s3fs-fuse/src/fdcache.cpp

1209 lines
32 KiB
C++
Raw Normal View History

/*
* s3fs - FUSE-based file system backed by Amazon S3
*
* Copyright 2007-2013 Takeshi Nakatani <ggtakec.com>
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
*/
#include <stdio.h>
#include <sys/stat.h>
#include <sys/types.h>
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
#include <sys/time.h>
#include <sys/file.h>
#include <unistd.h>
#include <pthread.h>
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
#include <syslog.h>
#include <errno.h>
#include <string.h>
#include <assert.h>
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
#include <curl/curl.h>
#include <openssl/crypto.h>
#include <string>
#include <iostream>
#include <sstream>
#include <map>
#include <list>
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
#include <vector>
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
#include "common.h"
#include "fdcache.h"
#include "s3fs.h"
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
#include "s3fs_util.h"
#include "curl.h"
using namespace std;
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
//------------------------------------------------
// Symbols
//------------------------------------------------
#define MAX_OBJECT_SIZE 68719476735LL // 64GB - 1L
#define MULTIPART_LOWLIMIT (20 * 1024 * 1024) // 20MB
#define FDPAGE_SIZE (50 * 1024 * 1024) // 50MB(parallel uploading is 5 parallel(default) * 10 MB)
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
//------------------------------------------------
// CacheFileStat class methods
//------------------------------------------------
bool CacheFileStat::MakeCacheFileStatPath(const char* path, string& sfile_path, bool is_create_dir)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
// make stat dir top path( "/<cache_dir>/.<bucket_name>.stat" )
string top_path = FdManager::GetCacheDir();
top_path += "/.";
top_path += bucket;
top_path += ".stat";
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(is_create_dir){
mkdirp(top_path + mydirname(path), 0777);
}
if(!path || '\0' == path[0]){
sfile_path = top_path;
}else{
sfile_path = top_path + SAFESTRPTR(path);
}
return true;
}
//------------------------------------------------
// CacheFileStat methods
//------------------------------------------------
CacheFileStat::CacheFileStat(const char* tpath) : path(""), fd(-1)
{
if(tpath && '\0' != tpath[0]){
SetPath(tpath, true);
}
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
CacheFileStat::~CacheFileStat()
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
Release();
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
bool CacheFileStat::SetPath(const char* tpath, bool is_open)
{
if(!tpath || '\0' == tpath[0]){
return false;
}
if(!Release()){
// could not close old stat file.
return false;
}
if(tpath){
path = tpath;
}
if(!is_open){
return true;
}
return Open();
}
bool CacheFileStat::Open(void)
{
if(0 == path.size()){
return false;
}
if(-1 != fd){
// already opened
return true;
}
// stat path
string sfile_path;
if(!CacheFileStat::MakeCacheFileStatPath(path.c_str(), sfile_path, true)){
FGPRINT("CacheFileStat::Open: failed to create cache stat file path(%s)\n", path.c_str());
SYSLOGERR("failed to create cache stat file path(%s)", path.c_str());
return false;
}
// open
if(-1 == (fd = open(sfile_path.c_str(), O_CREAT|O_RDWR, 0600))){
FGPRINT2("CacheFileStat::Open: failed to open cache stat file(%s) - errno(%d)\n", path.c_str(), errno);
//SYSLOGERR("failed to open cache stat file path(%s) - errno(%d)", path.c_str(), errno);
return false;
}
// lock
if(-1 == flock(fd, LOCK_EX)){
FGPRINT("CacheFileStat::Open: failed to lock cache stat file(%s) - errno(%d)\n", path.c_str(), errno);
SYSLOGERR("failed to lock cache stat file(%s) - errno(%d)", path.c_str(), errno);
close(fd);
fd = -1;
return false;
}
// seek top
if(0 != lseek(fd, 0, SEEK_SET)){
FGPRINT("CacheFileStat::Open: failed to lseek cache stat file(%s) - errno(%d)\n", path.c_str(), errno);
SYSLOGERR("failed to lseek cache stat file(%s) - errno(%d)", path.c_str(), errno);
flock(fd, LOCK_UN);
close(fd);
fd = -1;
return false;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
FGPRINT2(" CacheFileStat::Open: file locked(%s - %s)\n", path.c_str(), sfile_path.c_str());
return true;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
bool CacheFileStat::Release(void)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(-1 == fd){
// already release
return true;
}
// unlock
if(-1 == flock(fd, LOCK_UN)){
FGPRINT("CacheFileStat::Open: failed to unlock cache stat file(%s) - errno(%d)\n", path.c_str(), errno);
SYSLOGERR("failed to unlock cache stat file(%s) - errno(%d)", path.c_str(), errno);
return false;
}
FGPRINT2(" CacheFileStat::Open: file unlocked(%s)\n", path.c_str());
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(-1 == close(fd)){
FGPRINT("CacheFileStat::Open: failed to close cache stat file(%s) - errno(%d)\n", path.c_str(), errno);
SYSLOGERR("failed to close cache stat file(%s) - errno(%d)", path.c_str(), errno);
return false;
}
fd = -1;
return true;
}
//------------------------------------------------
// PageList methods
//------------------------------------------------
void PageList::FreeList(fdpage_list_t& list)
{
for(fdpage_list_t::iterator iter = list.begin(); iter != list.end(); iter = list.erase(iter)){
delete (*iter);
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
list.clear();
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
PageList::PageList(size_t size, bool is_init)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
Init(0, false);
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
PageList::~PageList()
{
Clear();
}
size_t PageList::Size(void) const
{
if(0 == pages.size()){
return 0;
}
fdpage_list_t::const_reverse_iterator riter = pages.rbegin();
return ((*riter)->offset + (*riter)->bytes);
}
int PageList::Resize(size_t size, bool is_init)
{
size_t total = Size();
if(0 == total){
Init(size, is_init);
}else if(total < size){
size_t remain = size - total; // remaining bytes
fdpage_list_t::reverse_iterator riter = pages.rbegin();
if((*riter)->bytes < FdManager::GetPageSize()){
// resize last area
remain += (*riter)->bytes; // remaining bytes(without last page)
(*riter)->bytes = remain > FdManager::GetPageSize() ? FdManager::GetPageSize() : remain; // reset page size
remain -= (*riter)->bytes; // remaining bytes(after last page)
(*riter)->init = is_init;
}
// add new area
for(off_t next = (*riter)->next(); 0 < remain; remain -= size, next += size){
size = remain > FdManager::GetPageSize() ? FdManager::GetPageSize() : remain;
fdpage* page = new fdpage(next, size, is_init);
pages.push_back(page);
}
}else if(total > size){
for(fdpage_list_t::reverse_iterator riter = pages.rbegin(); riter != pages.rend(); riter++){
if(static_cast<size_t>((*riter)->offset) < size){
(*riter)->bytes = size - (*riter)->offset;
break;
}
}
}
return true;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
void PageList::Clear(void)
{
PageList::FreeList(pages);
}
int PageList::Init(size_t size, bool is_init)
{
Clear();
for(size_t total = 0; total < size; total += FdManager::GetPageSize()){
size_t areasize = (total + FdManager::GetPageSize()) < size ? FdManager::GetPageSize() : (size - total);
fdpage* page = new fdpage(total, areasize, is_init);
pages.push_back(page);
}
return pages.size();
}
bool PageList::IsInit(off_t start, size_t size)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
off_t next = start + size;
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(0 == pages.size()){
return false;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
// check end
fdpage_list_t::reverse_iterator riter = pages.rbegin();
if((*riter)->next() < next){
// size is over end of page list.
return false;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
for(fdpage_list_t::iterator iter = pages.begin(); iter != pages.end(); iter++){
if(next <= (*iter)->offset){
break;
}
if((start <= (*iter)->offset && (*iter)->offset < next) || // start < iter-start < end
(start <= (*iter)->end() && (*iter)->end() < next) || // start < iter-end < end
((*iter)->offset <= start && next <= (*iter)->end()) ) // iter-start < start < end < iter-end
{
if(!(*iter)->init){
return false;
}
}
}
return true;
}
bool PageList::SetInit(off_t start, size_t size, bool is_init)
{
// check size & resize
if(Size() < (start + size)){
Resize(start + size, false);
}
off_t next = start + size;
for(fdpage_list_t::iterator iter = pages.begin(); iter != pages.end(); iter++){
if((*iter)->end() < start){
// out of area
// iter:start < iter:end < start < end
continue;
}else if(next <= (*iter)->offset){
// out of area
// start < end < iter:start < iter:end
break;
}
// area of target overlaps with iter area
// iter:start < start < iter:end < end
// iter:start < start < end < iter:end
// start < iter:start < iter:end < end
// start < iter:start < end < iter:end
if((*iter)->init != is_init){
(*iter)->init = is_init;
}
}
return true;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
bool PageList::FindUninitPage(off_t start, off_t& resstart, size_t& ressize)
{
for(fdpage_list_t::iterator iter = pages.begin(); iter != pages.end(); iter++){
if(start <= (*iter)->end()){
if(!(*iter)->init){
resstart = (*iter)->offset;
ressize = (*iter)->bytes;
return true;
}
}
}
return false;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
int PageList::GetUninitPages(fdpage_list_t& uninit_list, off_t start)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
for(fdpage_list_t::iterator iter = pages.begin(); iter != pages.end(); iter++){
if(start <= (*iter)->end()){
// after start pos
if(!(*iter)->init){
// found uninitialized area
fdpage_list_t::reverse_iterator riter = uninit_list.rbegin();
if(riter != uninit_list.rend() && (*riter)->next() == (*iter)->offset){
// merge to before page
(*riter)->bytes += (*iter)->bytes;
}else{
fdpage* page = new fdpage((*iter)->offset, (*iter)->bytes, false);
uninit_list.push_back(page);
}
}
}
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
return uninit_list.size();
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
bool PageList::Serialize(CacheFileStat& file, bool is_output)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(!file.Open()){
return false;
}
if(is_output){
//
// put to file
//
stringstream ssall;
ssall << Size();
for(fdpage_list_t::iterator iter = pages.begin(); iter != pages.end(); iter++){
ssall << "\n" << (*iter)->offset << ":" << (*iter)->bytes << ":" << ((*iter)->init ? "1" : "0");
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
string strall = ssall.str();
if(0 >= pwrite(file.GetFd(), strall.c_str(), strall.length(), 0)){
FGPRINT("PageList::Serialize: failed to write stats(%d)\n", errno);
SYSLOGERR("failed to write stats(%d)", errno);
return false;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
}else{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
//
// loading from file
//
struct stat st;
memset(&st, 0, sizeof(struct stat));
if(-1 == fstat(file.GetFd(), &st)){
FGPRINT("PageList::Serialize: fstat is failed. errno(%d)\n", errno);
SYSLOGERR("fstat is failed. errno(%d)", errno);
return false;
}
if(0 >= st.st_size){
// nothing
Init(0, false);
return true;
}
char* ptmp;
if(NULL == (ptmp = (char*)calloc(st.st_size + 1, sizeof(char)))){
FGPRINT("PageList::Serialize: could not allocate memory.\n");
SYSLOGERR("could not allocate memory.");
S3FS_FUSE_EXIT();
return false;
}
// read from file
if(0 >= pread(file.GetFd(), ptmp, st.st_size, 0)){
FGPRINT("PageList::Serialize: failed to read stats(%d)\n", errno);
SYSLOGERR("failed to read stats(%d)", errno);
free(ptmp);
return false;
}
string oneline;
stringstream ssall(ptmp);
free(ptmp);
// init
Clear();
// load(size)
if(!getline(ssall, oneline, '\n')){
FGPRINT("PageList::Serialize: failed to parse stats.\n");
SYSLOGERR("failed to parse stats.");
return false;
}
size_t total = static_cast<size_t>(atoi(oneline.c_str()));
// load each part
bool is_err = false;
while(getline(ssall, oneline, '\n')){
string part;
stringstream ssparts(oneline);
// offset
if(!getline(ssparts, part, ':')){
is_err = true;
break;
}
off_t offset = static_cast<off_t>(atoi(part.c_str()));
// size
if(!getline(ssparts, part, ':')){
is_err = true;
break;
}
ssize_t size = static_cast<ssize_t>(atoi(part.c_str()));
// init
if(!getline(ssparts, part, ':')){
is_err = true;
break;
}
bool is_init = (1 == atoi(part.c_str()) ? true : false);
// add new area
SetInit(offset, size, is_init);
}
if(is_err){
FGPRINT("PageList::Serialize: failed to parse stats.\n");
SYSLOGERR("failed to parse stats.");
Clear();
return false;
}
// check size
if(total != Size()){
FGPRINT("PageList::Serialize: different size(%zd - %zd).\n", total, Size());
SYSLOGERR("different size(%zd - %zd).", total, Size());
Clear();
return false;
}
}
return true;
}
void PageList::Dump(void)
{
int cnt = 0;
FGPRINT2(" FdEntity::Dump = {\n");
for(fdpage_list_t::iterator iter = pages.begin(); iter != pages.end(); iter++, cnt++){
FGPRINT2(" [%08d] -> {%014zd - %014zd : %s}\n", cnt, (*iter)->offset, (*iter)->bytes, (*iter)->init ? "true" : "false");
}
FGPRINT2(" }\n");
}
//------------------------------------------------
// FdEntity methods
//------------------------------------------------
FdEntity::FdEntity(const char* tpath, const char* cpath)
: is_lock_init(false), path(SAFESTRPTR(tpath)), cachepath(SAFESTRPTR(cpath)), fd(-1), file(NULL), is_modify(false)
{
try{
pthread_mutex_init(&fdent_lock, NULL);
is_lock_init = true;
}catch(exception& e){
FGPRINT("FdEntity::FdEntity: failed to init mutex\n");
SYSLOGERR("failed to init mutex");
}
}
FdEntity::~FdEntity()
{
Clear();
if(is_lock_init){
try{
pthread_mutex_destroy(&fdent_lock);
}catch(exception& e){
FGPRINT("FdEntity::~FdEntity: failed to destroy mutex\n");
SYSLOGERR("failed to destroy mutex");
}
is_lock_init = false;
}
}
void FdEntity::Clear(void)
{
AutoLock auto_lock(&fdent_lock);
if(file){
{
CacheFileStat cfstat(path.c_str());
if(!pagelist.Serialize(cfstat, true)){
FGPRINT("FdEntity::Clear: failed to save cache stat file(%s).\n", path.c_str());
SYSLOGERR("failed to save cache stat file(%s).", path.c_str());
}
}
fclose(file);
file = NULL;
fd = -1;
}
pagelist.Init(0, false);
refcnt = 0;
path = "";
cachepath = "";
is_modify = false;
}
void FdEntity::Close(void)
{
FGPRINT2(" FdEntity::Close[path=%s][fd=%d][refcnt=%d]\n", path.c_str(), fd, (-1 != fd ? refcnt - 1 : refcnt));
if(-1 != fd){
AutoLock auto_lock(&fdent_lock);
if(0 < refcnt){
refcnt--;
}
if(0 == refcnt){
{
CacheFileStat cfstat(path.c_str());
if(!pagelist.Serialize(cfstat, true)){
FGPRINT("FdEntity::Close: failed to save cache stat file(%s).\n", path.c_str());
SYSLOGERR("failed to save cache stat file(%s).", path.c_str());
}
}
fclose(file);
file = NULL;
fd = -1;
}
}
}
int FdEntity::Dup(void)
{
FGPRINT2(" FdEntity::Dup[path=%s][fd=%d][refcnt=%d]\n", path.c_str(), fd, (-1 != fd ? refcnt + 1 : refcnt));
if(-1 != fd){
AutoLock auto_lock(&fdent_lock);
refcnt++;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
return fd;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
int FdEntity::Open(ssize_t size, time_t time)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
bool already_opened = false; // already opened fd
bool is_csf_loaded = false; // loaded by cache stat file
bool is_truncate = false; // need to truncate
bool init_value = false; // value for pagelist
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
FGPRINT2(" FdEntity::Open[path=%s][fd=%d][size=%zd][time=%zd]\n", path.c_str(), fd, size, time);
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(-1 != fd){
// already opened, needs to increment refcnt.
already_opened = true;
}else{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
// open
if(0 != cachepath.size()){
// At first, open & flock stat file.
{
CacheFileStat cfstat(path.c_str());
is_csf_loaded = pagelist.Serialize(cfstat, false);
}
// open cache file
if(is_csf_loaded && -1 != (fd = open(cachepath.c_str(), O_RDWR))){
// file exists
struct stat st;
memset(&st, 0, sizeof(struct stat));
if(-1 == fstat(fd, &st)){
FGPRINT("FdEntity::Open: fstat is failed. errno(%d)\n", errno);
SYSLOGERR("fstat is failed. errno(%d)", errno);
fclose(file);
file = NULL;
fd = -1;
return (0 == errno ? -EIO : -errno);
}
if(static_cast<size_t>(st.st_size) != pagelist.Size()){
is_csf_loaded = false; // reinitializing
if(-1 == size){
size = st.st_size;
}else{
is_truncate = true;
}
}else{
// size OK! --> no initialize after this line.
}
}else{
// file does not exist -> create & open
if(-1 == (fd = open(cachepath.c_str(), O_CREAT|O_RDWR|O_TRUNC, 0600))){
FGPRINT("FdEntity::Open: failed to open file(%s). errno(%d)\n", cachepath.c_str(), errno);
SYSLOGERR("failed to open file(%s). errno(%d)", cachepath.c_str(), errno);
return (0 == errno ? -EIO : -errno);
}
if(-1 == size){
size = 0;
}else{
is_truncate = true;
}
is_csf_loaded = false;
}
// make file pointer(for being same tmpfile)
if(NULL == (file = fdopen(fd, "wb"))){
FGPRINT("FdEntity::Open: failed to get fileno(%s). errno(%d)\n", cachepath.c_str(), errno);
SYSLOGERR("failed to get fileno(%s). errno(%d)", cachepath.c_str(), errno);
close(fd);
fd = -1;
return (0 == errno ? -EIO : -errno);
}
}else{
// open temporary file
if(NULL == (file = tmpfile()) || -1 ==(fd = fileno(file))){
FGPRINT("FdEntity::Open: failed to open tmp file. err(%d)\n", errno);
SYSLOGERR("failed to open tmp file. err(%d)", errno);
if(file){
fclose(file);
file = NULL;
}
return (0 == errno ? -EIO : -errno);
}
if(-1 == size){
size = 0;
}else{
is_truncate = true;
}
}
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
// truncate
if(is_truncate){
if(0 != ftruncate(fd, size) || 0 != fsync(fd)){
FGPRINT("FdEntity::Open: ftruncate(%s) or fsync returned err(%d)\n", cachepath.c_str(), errno);
SYSLOGERR("ftruncate(%s) or fsync returned err(%d)", cachepath.c_str(), errno);
fclose(file);
file = NULL;
fd = -1;
return (0 == errno ? -EIO : -errno);
}
}
// set mtime
if(-1 != time){
if(0 != SetMtime(time)){
FGPRINT("FdEntity::Open: failed to set mtime. errno(%d)\n", errno);
SYSLOGERR("failed to set mtime. errno(%d)", errno);
fclose(file);
file = NULL;
fd = -1;
return (0 == errno ? -EIO : -errno);
}
}
// set internal data
if(already_opened){
Dup();
}else{
if(!is_csf_loaded){
pagelist.Init(size, init_value);
}
refcnt = 1;
is_modify = false;
}
return 0;
}
int FdEntity::SetMtime(time_t time)
{
FGPRINT2(" FdEntity::SetMtime[path=%s][fd=%d][time=%zd]\n", path.c_str(), fd, time);
if(-1 == time){
return 0;
}
if(-1 != fd){
AutoLock auto_lock(&fdent_lock);
struct timeval tv[2];
tv[0].tv_sec = time;
tv[0].tv_usec= 0L;
tv[1].tv_sec = tv[0].tv_sec;
tv[1].tv_usec= 0L;
if(-1 == futimes(fd, tv)){
FGPRINT("FdEntity::Set: futimes failed. errno(%d)\n", errno);
SYSLOGERR("futimes failed. errno(%d)", errno);
return -errno;
}
}else if(0 < cachepath.size()){
// not opened file yet.
struct utimbuf n_mtime;
n_mtime.modtime = time;
n_mtime.actime = time;
if(-1 == utime(cachepath.c_str(), &n_mtime)){
//FGPRINT("FdEntity::Set: utime failed. errno(%d)\n", errno);
//SYSLOGERR("utime failed. errno(%d)", errno);
return -errno;
}
}
return 0;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
bool FdEntity::GetSize(size_t& size)
{
if(-1 == fd){
return false;
}
AutoLock auto_lock(&fdent_lock);
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
size = pagelist.Size();
return true;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
bool FdEntity::GetMtime(time_t& time)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
struct stat st;
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(!GetStats(st)){
return false;
}
time = st.st_mtime;
return true;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
bool FdEntity::GetStats(struct stat& st)
{
if(-1 == fd){
return false;
}
AutoLock auto_lock(&fdent_lock);
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
memset(&st, 0, sizeof(struct stat));
if(-1 == fstat(fd, &st)){
FGPRINT("FdEntity::GetStats: fstat failed. errno(%d)\n", errno);
SYSLOGERR("fstat failed. errno(%d)", errno);
return false;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
return true;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
bool FdEntity::SetAllStatus(bool is_enable)
{
FGPRINT2(" FdEntity::SetAllStatus[path=%s][fd=%d][%s]\n", path.c_str(), fd, is_enable ? "enable" : "disable");
if(-1 == fd){
return false;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
AutoLock auto_lock(&fdent_lock);
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
// get file size
struct stat st;
memset(&st, 0, sizeof(struct stat));
if(-1 == fstat(fd, &st)){
FGPRINT("FdEntity::SetAllEnable: fstat is failed. errno(%d)\n", errno);
SYSLOGERR("fstat is failed. errno(%d)", errno);
return false;
}
// Reinit
pagelist.Init(st.st_size, is_enable);
return true;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
int FdEntity::Load(off_t start, ssize_t size)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
int result = 0;
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
FGPRINT2(" FdEntity::Load[path=%s][fd=%d][offset=%zd][size=%zd]\n", path.c_str(), fd, start, size);
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(-1 == fd){
return -EBADF;
}
AutoLock auto_lock(&fdent_lock);
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
// check loaded area & load
fdpage_list_t uninit_list;
if(0 < pagelist.GetUninitPages(uninit_list, start)){
for(fdpage_list_t::iterator iter = uninit_list.begin(); iter != uninit_list.end(); iter++){
if(-1 != size && (start + size) <= (*iter)->offset){
break;
}
// download
if((*iter)->bytes >= MULTIPART_LOWLIMIT && !nomultipart){ // 20MB
// parallel request
// Additional time is needed for large files
time_t backup = 0;
if(120 > S3fsCurl::GetReadwriteTimeout()){
backup = S3fsCurl::SetReadwriteTimeout(120);
}
result = S3fsCurl::ParallelGetObjectRequest(path.c_str(), fd, (*iter)->offset, (*iter)->bytes);
if(0 != backup){
S3fsCurl::SetReadwriteTimeout(backup);
}
}else{
// single request
S3fsCurl s3fscurl;
result = s3fscurl.GetObjectRequest(path.c_str(), fd, (*iter)->offset, (*iter)->bytes);
}
if(0 != result){
break;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
// Set init flag
pagelist.SetInit((*iter)->offset, (*iter)->bytes, true);
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
PageList::FreeList(uninit_list);
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
return result;
}
bool FdEntity::LoadFull(size_t* size, bool force_load)
{
int result;
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
FGPRINT2(" FdEntity::LoadFull[path=%s][fd=%d]\n", path.c_str(), fd);
if(-1 == fd){
if(0 != Open()){
return false;
}
}
if(force_load){
SetAllDisable();
}
//
// TODO: possibly do background for delay loading
//
if(0 != (result = Load(0, pagelist.Size()))){
FGPRINT("FdEntity::FullDownload: could not download, result(%d)\n", result);
SYSLOGERR("could not download, result(%d)", result);
return false;
}
if(is_modify){
AutoLock auto_lock(&fdent_lock);
is_modify = false;
}
if(size){
*size = pagelist.Size();
}
return true;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
int FdEntity::RowFlush(const char* tpath, headers_t& meta, bool ow_sse_flg, bool force_sync)
{
int result;
FGPRINT2(" FdEntity::RowFlush[tpath=%s][path=%s][fd=%d]\n", SAFESTRPTR(tpath), path.c_str(), fd);
if(-1 == fd){
return -EBADF;
}
AutoLock auto_lock(&fdent_lock);
if(!force_sync && !is_modify){
// nothing to update.
return 0;
}
/*
* Make decision to do multi upload (or not) based upon file size
*
* According to the AWS spec:
* - 1 to 10,000 parts are allowed
* - minimum size of parts is 5MB (expect for the last part)
*
* For our application, we will define part size to be 10MB (10 * 2^20 Bytes)
* maximum file size will be ~64 GB - 2 ** 36
*
* Initially uploads will be done serially
*
* If file is > 20MB, then multipart will kick in
*/
if(pagelist.Size() > MAX_OBJECT_SIZE){ // 64GB - 1
// close f ?
return -ENOTSUP;
}
// seek to head of file.
if(0 != lseek(fd, 0, SEEK_SET)){
FGPRINT("FdEntity::RowFlush: lseek error(%d)\n", errno);
SYSLOGERR("lseek error(%d)\n", errno);
return -errno;
}
if(pagelist.Size() >= MULTIPART_LOWLIMIT && !nomultipart){ // 20MB
// Additional time is needed for large files
time_t backup = 0;
if(120 > S3fsCurl::GetReadwriteTimeout()){
backup = S3fsCurl::SetReadwriteTimeout(120);
}
result = S3fsCurl::ParallelMultipartUploadRequest(tpath ? tpath : path.c_str(), meta, fd, ow_sse_flg);
if(0 != backup){
S3fsCurl::SetReadwriteTimeout(backup);
}
}else{
S3fsCurl s3fscurl;
result = s3fscurl.PutRequest(tpath ? tpath : path.c_str(), meta, fd, ow_sse_flg);
}
// seek to head of file.
if(0 == result && 0 != lseek(fd, 0, SEEK_SET)){
FGPRINT("FdEntity::RowFlush: lseek error(%d)\n", errno);
SYSLOGERR("lseek error(%d)\n", errno);
return -errno;
}
if(0 == result){
is_modify = false;
}
return result;
}
ssize_t FdEntity::Read(char* bytes, off_t start, size_t size, bool force_load)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
int result;
ssize_t rsize;
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
FGPRINT2(" FdEntity::Read[path=%s][fd=%d][offset=%zd][size=%zd]\n", path.c_str(), fd, start, size);
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(-1 == fd){
return -EBADF;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(force_load){
AutoLock auto_lock(&fdent_lock);
pagelist.SetInit(start, size, false);
}
// Loading
if(0 != (result = Load(start, size))){
FGPRINT("FdEntity::Read: could not download. start(%zd), size(%zd), errno(%d)\n", start, size, result);
SYSLOGERR("could not download. start(%zd), size(%zd), errno(%d)", start, size, result);
return -EIO;
}
// Reading
{
AutoLock auto_lock(&fdent_lock);
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(-1 == (rsize = pread(fd, bytes, size, start))){
FGPRINT("FdEntity::Read: pread failed. errno(%d)\n", errno);
SYSLOGERR("pread failed. errno(%d)", errno);
return -errno;
}
}
return rsize;
}
ssize_t FdEntity::Write(const char* bytes, off_t start, size_t size)
{
int result;
ssize_t wsize;
FGPRINT2(" FdEntity::Write[path=%s][fd=%d][offset=%zd][size=%zd]\n", path.c_str(), fd, start, size);
if(-1 == fd){
return -EBADF;
}
// Load unitialized area which starts from 0 to (start + size) before writing.
if(0 != (result = Load(0, start))){
FGPRINT("FdEntity::Write: failed to load uninitialized area before writing(errno=%d)\n", result);
SYSLOGERR("failed to load uninitialized area before writing(errno=%d)", result);
return static_cast<ssize_t>(result);
}
// Writing
{
AutoLock auto_lock(&fdent_lock);
if(-1 == (wsize = pwrite(fd, bytes, size, start))){
FGPRINT("FdEntity::Write: pwrite failed. errno(%d)\n", errno);
SYSLOGERR("pwrite failed. errno(%d)", errno);
return -errno;
}
if(!is_modify){
is_modify = true;
}
if(0 < wsize){
pagelist.SetInit(start, wsize, true);
}
}
return wsize;
}
//------------------------------------------------
// FdManager class valiable
//------------------------------------------------
FdManager FdManager::singleton;
pthread_mutex_t FdManager::fd_manager_lock;
bool FdManager::is_lock_init(false);
string FdManager::cache_dir("");
size_t FdManager::page_size(FDPAGE_SIZE);
//------------------------------------------------
// FdManager class methods
//------------------------------------------------
bool FdManager::SetCacheDir(const char* dir)
{
if(!dir || '\0' == dir[0]){
cache_dir = "";
}else{
cache_dir = dir;
}
return true;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
size_t FdManager::SetPageSize(size_t size)
{
// If already has entries, this function is failed.
if(0 < FdManager::get()->fent.size()){
return -1;
}
size_t old = FdManager::page_size;
FdManager::page_size = size;
return old;
}
bool FdManager::DeleteCacheDirectory(void)
{
if(0 == FdManager::cache_dir.size()){
return true;
}
string cache_dir;
if(!FdManager::MakeCachePath(NULL, cache_dir, false)){
return false;
}
return delete_files_in_dir(cache_dir.c_str(), true);
}
int FdManager::DeleteCacheFile(const char* path)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
FGPRINT2(" FdManager::DeleteCacheFile[path=%s]\n", SAFESTRPTR(path));
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
if(!path){
return -EIO;
}
if(0 == FdManager::cache_dir.size()){
return 0;
}
string cache_path = "";
if(!FdManager::MakeCachePath(path, cache_path, false)){
return 0;
}
if(0 != unlink(cache_path.c_str())){
//FGPRINT("FdManager::DeleteCacheFile: failed to delete file(%s): errno=%d\n", path, errno);
//SYSLOGERR("failed to delete file(%s): errno=%d", path, errno);
return -errno;
}
return 0;
}
bool FdManager::MakeCachePath(const char* path, string& cache_path, bool is_create_dir)
{
if(0 == FdManager::cache_dir.size()){
cache_path = "";
return true;
}
string resolved_path(FdManager::cache_dir + "/" + bucket);
if(is_create_dir){
mkdirp(resolved_path + mydirname(path), 0777);
}
if(!path || '\0' == path[0]){
cache_path = resolved_path;
}else{
cache_path = resolved_path + SAFESTRPTR(path);
}
return true;
}
//------------------------------------------------
// FdManager methods
//------------------------------------------------
FdManager::FdManager()
{
if(this == FdManager::get()){
try{
pthread_mutex_init(&FdManager::fd_manager_lock, NULL);
FdManager::is_lock_init = true;
}catch(exception& e){
FdManager::is_lock_init = false;
FGPRINT("FdManager::FdManager: failed to init mutex\n");
SYSLOGERR("failed to init mutex");
}
}else{
assert(false);
}
}
FdManager::~FdManager()
{
if(this == FdManager::get()){
for(fdent_map_t::iterator iter = fent.begin(); fent.end() != iter; iter++){
FdEntity* ent = (*iter).second;
delete ent;
}
fent.clear();
if(FdManager::is_lock_init){
try{
pthread_mutex_destroy(&FdManager::fd_manager_lock);
}catch(exception& e){
FGPRINT("FdManager::FdManager: failed to init mutex\n");
SYSLOGERR("failed to init mutex");
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
FdManager::is_lock_init = false;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
}else{
assert(false);
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
FdEntity* FdManager::GetFdEntity(const char* path)
{
FGPRINT2(" FdManager::GetFdEntity[path=%s]\n", SAFESTRPTR(path));
if(!path || '\0' == path[0]){
return NULL;
}
AutoLock auto_lock(&FdManager::fd_manager_lock);
fdent_map_t::iterator iter = fent.find(string(path));
if(fent.end() == iter){
return NULL;
}
return (*iter).second;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
FdEntity* FdManager::Open(const char* path, ssize_t size, time_t time, bool force_tmpfile, bool is_create)
{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
FdEntity* ent;
FGPRINT2(" FdManager::Open[path=%s][size=%zd][time=%zd]\n", SAFESTRPTR(path), size, time);
if(!path || '\0' == path[0]){
return NULL;
}
AutoLock auto_lock(&FdManager::fd_manager_lock);
fdent_map_t::iterator iter = fent.find(string(path));
if(fent.end() != iter){
// found
ent = (*iter).second;
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
}else if(is_create){
// not found
string cache_path = "";
if(!force_tmpfile && !FdManager::MakeCachePath(path, cache_path, true)){
FGPRINT("FdManager::GetFd: failed to make cache path for object(%s).\n", path);
SYSLOGERR("failed to make cache path for object(%s).", path);
return NULL;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
// make new obj
ent = new FdEntity(path, cache_path.c_str());
fent[string(path)] = ent;
}else{
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
return NULL;
}
Changes codes for performance(part 3) * Summay This revision includes big change about temporary file and local cache file. By this big change, s3fs works with good performance when s3fs opens/ closes/syncs/reads object. I made a big change about the handling about temporary file and local cache file to do this implementation. * Detail 1) About temporary file(local file) s3fs uses a temporary file on local file system when s3fs does download/ upload/open/seek object on S3. After this revision, s3fs calls ftruncate() function when s3fs makes the temporary file. In this way s3fs can set a file size of precisely length without downloading. (Notice - ftruncate function is for XSI-compliant systems, so that possibly you have a problem on non-XSI-compliant systems.) By this change, s3fs can download a part of a object by requesting with "Range" http header. It seems like downloading by each block unit. The default block(part) size is 50MB, it is caused the result which is default parallel requests count(5) by default multipart upload size(10MB). If you need to change this block size, you can change by new option "fd_page_size". This option can take from 1MB(1024 * 1024) to any bytes. So that, you have to take care about that fdcache.cpp(and fdcache.h) were changed a lot. 2) About local cache Local cache files which are in directory specified by "use_cache" option do not have always all of object data. This cause is that s3fs uses ftruncate function and reads(writes) each block unit of a temporary file. s3fs manages each block unit's status which are "downloaded area" or "not". For this status, s3fs makes new temporary file in cache directory which is specified by "use_cache" option. This status files is in a directory which is named "<use_cache sirectory>/.<bucket_name>/". When s3fs opens this status file, s3fs locks this file for exclusive control by calling flock function. You need to take care about this, the status files can not be laid on network drive(like NFS). This revision changes about file open mode, s3fs always opens a local cache file and each status file with writable mode. Last, this revision adds new option "del_cache", this option means that s3fs deletes all local cache file when s3fs starts and exits. 3) Uploading When s3fs writes data to file descriptor through FUSE request, old s3fs revision downloads all of the object. But new revision does not download all, it downloads only small percial area(some block units) including writing data area. And when s3fs closes or flushes the file descriptor, s3fs downloads other area which is not downloaded from server. After that, s3fs uploads all of data. Already r456 revision has parallel upload function, then this revision with r456 and r457 are very big change for performance. 4) Downloading By changing a temporary file and a local cache file, when s3fs downloads a object, it downloads only the required range(some block units). And s3fs downloads units by parallel GET request, it is same as a case of uploading. (Maximum parallel request count and each download size are specified same parameters for uploading.) In the new revision, when s3fs opens file, s3fs returns file descriptor soon. Because s3fs only opens(makes) the file descriptor with no downloading data. And when s3fs reads a data, s3fs downloads only some block unit including specified area. This result is good for performance. 5) Changes option name The option "parallel_upload" which added at r456 is changed to new option name as "parallel_count". This reason is this option value is not only used by uploading object, but a uploading object also uses this option. (For a while, you can use old option name "parallel_upload" for compatibility.) git-svn-id: http://s3fs.googlecode.com/svn/trunk@458 df820570-a93a-0410-bd06-b72b767a4274
2013-07-23 16:01:48 +00:00
// open
if(-1 == ent->Open(size, time)){
return NULL;
}
return ent;
}
bool FdManager::Close(FdEntity* ent)
{
FGPRINT2(" FdManager::Close[ent->file=%s][ent->fd=%d]\n", ent ? ent->GetPath() : "", ent ? ent->GetFd() : -1);
AutoLock auto_lock(&FdManager::fd_manager_lock);
for(fdent_map_t::iterator iter = fent.begin(); iter != fent.end(); iter++){
if((*iter).second == ent){
ent->Close();
if(!ent->IsOpen()){
delete (*iter).second;
fent.erase(iter);
return true;
}
}
}
return false;
}