2
2
mirror of https://github.com/octoleo/restic.git synced 2024-12-23 11:28:54 +00:00
Commit Graph

696 Commits

Author SHA1 Message Date
Michael Eischer
dcf9ded977 dump: Use slashes as path separators in tar on windows
This conversion is necessary as paths returned by filepath.Rel use the
operating-system native path separator
2020-08-30 18:26:50 +02:00
Simon Beck
fe6445e0f4 dump: Added unit tests for tar output 2020-08-30 16:00:48 +02:00
Simon Beck
80a11960dd dump: Always dump relative paths into tarballs
Tarballs should only contain relative paths.
2020-08-29 21:14:34 +02:00
Simon Beck
d6f739ec22 dump: Extract tar code to internal/dump 2020-08-29 21:14:34 +02:00
Michael Eischer
b25978a53c backup: Fix reporting of directory count in summary
Previously the directory stats were reported immediately after calling
`SaveDir`. However, as the latter method saves the tree asynchronously
the stats were still initialized to their nil value. The stats are now
reported via a callback similar to the one used for the fileSaver.
2020-08-27 22:43:51 +02:00
Alexander Neumann
b1b3f1ecb6
Merge pull request #2674 from MichaelEischer/prune-strict-checks
prune: Stricter error checks
2020-08-23 10:30:41 +02:00
Hristo Trendev
51c22f4223 local backend: ignore not supported error on sync()
Closes #2395
2020-08-22 01:27:07 +02:00
Michael Eischer
f4b9544ab2 prune: Add test that repack aborts on wrong blob 2020-08-16 11:34:01 +02:00
Michael Eischer
367449dede prune: Reduce memory allocations while repacking
The slicing operator `slice[low:high]` default to 0 for the lower bound and
len(slice) for the upper bound when either or both are not specified.
Fix the code to use `cap(slice)` to check for the slice capacity.
2020-08-16 11:34:01 +02:00
Michael Eischer
7042bafea5 prune: Abort repacking when a pack contains a wrong blob
If a blob in a pack file can be decrypted successfully but contains data
that results in a different hash than stated in the header pack, then
abort repacking. As both the pack header and the blob are
cryptographically verified this either means than a malicious entity
tampered with the backup or indicates hardware problems on the client.
prune should fail with an error in both cases.
2020-08-16 11:34:01 +02:00
aawsome
0fed6a8dfc
Use "pack file" instead of "data file" (#2885)
- changed variable names, especially changed DataFile into PackFile
- changed in some comments
- always use "pack file" in docu
2020-08-16 11:16:38 +02:00
MichaelEischer
eca0f0ad24
Merge pull request #2863 from aawsome/index-no-duplicates
Don't save exact duplicates in merged index
2020-08-08 18:24:14 +02:00
Alexander Weiss
b112533812 Don't save exact duplicates when merging indexes 2020-08-05 06:32:02 +02:00
Alexander Weiss
5e63294355 Add benchmark MasterIndexAlloc 2020-08-05 06:32:02 +02:00
Michael Eischer
05116e4787 prune: Cleanup progress bar handling while repacking 2020-08-03 19:32:46 +02:00
Alexander Neumann
b7b479b668
Merge pull request #2599 from MichaelEischer/tweak-mem-usage
Reduce memory usage when searching for used blobs
2020-08-01 13:49:02 +02:00
Alexander Neumann
4cf9656f12
Merge pull request #2861 from MichaelEischer/fix-rclone-crash
rclone: Don't panic after unexpected subprocess exit
2020-08-01 13:08:50 +02:00
Alexander Neumann
2580eef2aa
Merge pull request #2318 from classmarkets/2175-named-keys
Allow specifying user and host when adding keys
2020-08-01 13:06:31 +02:00
Michael Eischer
a178e5628e Remove duplicate TreeLoader interface 2020-08-01 12:29:16 +02:00
Michael Eischer
af66a62c04 FindUsedBlobs: Test that seen blobs are skipped
This also copies the TreeLoader interface from internal/walker to allow
stubbing the repository in the call to `FindUsedBlobs`.
2020-08-01 12:29:16 +02:00
Michael Eischer
9ea1a78bd4 FindUsedBlobs: Check for seen blobs before loading trees
The only effective change in behavior is that that toplevel nodes can
also be skipped.
2020-08-01 12:29:16 +02:00
Michael Eischer
184103647a FindUsedBlobs: merge seen into blobs BlobSet
The seen BlobSet always contained a subset of the entries in blobs.
Thus use blobs instead and avoid the memory overhead of the second set.

Suggested-by: Alexander Weiss <alex@weissfam.de>
2020-08-01 12:29:16 +02:00
Michael Eischer
c81b122374 rclone: Don't panic after unexpected subprocess exit
As the connection to the rclone child process is now closed after an
unexpected subprocess exit, later requests will cause the http2
transport to try to reestablish a new connection. As previously this never
should have happened, the connection called panic in that case. This
panic is now replaced with a simple error message, as it no longer
indicates an internal problem.
2020-08-01 12:17:40 +02:00
Alexander Weiss
f8316948d1 Optimize FUSE - make command restic mount faster and consume less memory
- Add Open() functionality to dir
- only access index for blobs when file is read
- Implement NodeOpener and put one-time file stuff there
- Add comment about locking as suggested by bazil.org/fuse

=> Thanks at Michael Eischer for suggesting the last two improvements
2020-07-28 23:01:18 +02:00
Michael Eischer
ea97ff1ba4 rclone: Skip crash test when rclone is not found 2020-07-26 12:06:18 +02:00
Michael Eischer
01b9581453 rclone: Better field names for stdio conn 2020-07-26 00:29:25 +02:00
Michael Eischer
3cd927d180 rclone: Give rclone time to finish before closing stdin pipe
Calling `Close()` on the rclone backend sometimes failed during test
execution with 'signal: Broken pipe'. The stdio connection closed both
the stdin and stdout file descriptors at the same moment, therefore
giving rclone no chance to properly send any final http2 data frames.

Now the stdin connection to rclone is closed first and will only be
forcefully closed after a timeout. In case rclone exits before the
timeout then the stdio connection will be closed normally.
2020-07-26 00:29:25 +02:00
greatroar
bf7b1f12ea rclone: Add test for pipe handling when rclone exits 2020-07-26 00:29:25 +02:00
Michael Eischer
8554332894 rclone: Close rclone side of stdio_conn pipes
restic did not notice when the rclone subprocess exited unexpectedly.

restic manually created pipes for stdin and stdout and used these for the
connection to the rclone subprocess. The process creating a pipe gets
file descriptors for the sender and receiver side of a pipe and passes
them on to the subprocess. The expected behavior would be that reads or
writes in the parent process fail / return once the child process dies
as a pipe would now just have a reader or writer but not both.

However, this never happened as restic kept the reader and writer
file descriptors of the pipes. `cmd.StdinPipe` and `cmd.StdoutPipe`
close the subprocess side of pipes once the child process was started
and close the parent process side of pipes once wait has finished. We
can't use these functions as we need access to the raw `os.File` so just
replicate that behavior.
2020-07-26 00:29:25 +02:00
greatroar
3e93b36ca4 Make rclone.New private 2020-07-26 00:28:45 +02:00
Michael Eischer
c847aace35 Rename Index interface to MasterIndex
The interface is now only implemented by repository.MasterIndex.
2020-07-25 21:19:46 +02:00
Alexander Weiss
9d1fb94c6c make Lookup() return all blobs
+ simplify syntax
2020-07-25 21:18:34 +02:00
MichaelEischer
020cab8e08
Merge pull request #2787 from greatroar/no-blobsize-cache
Remove blob size cache from restic mount
2020-07-25 20:46:35 +02:00
greatroar
07da61baee Remove blob size cache from restic mount 2020-07-25 19:39:14 +02:00
MichaelEischer
3d530dfc91
Merge pull request #2827 from aawsome/archiver-test-contents
Make self-healing work when backing up with parent snapshot
2020-07-25 13:13:18 +02:00
MichaelEischer
bbc960f957
Merge pull request #2635 from greatroar/optimize-sortbycached
Optimize sorting blobs by cache status
2020-07-25 12:35:42 +02:00
greatroar
309598c237 Simplify sortCachedPacksFirst test in internal/repository
The test now uses the fact that the sort is stable. It's not guaranteed
to be, but the test is cleaner and more exhaustive. sortCachedPacksFirst
no longer needs a return value.
2020-07-25 12:12:59 +02:00
greatroar
03d23e6faa Speed up blob sorting in internal/repository
name                    old time/op    new time/op    delta
SortCachedPacksFirst-8     208µs ± 3%     186µs ± 3%  -10.74%  (p=0.000 n=10+8)

name                    old alloc/op   new alloc/op   delta
SortCachedPacksFirst-8     213kB ± 0%     139kB ± 0%  -34.62%  (p=0.000 n=10+10)

name                    old allocs/op  new allocs/op  delta
SortCachedPacksFirst-8     1.03k ± 0%     1.03k ± 0%   -0.19%  (p=0.000 n=10+10)
2020-07-25 12:12:59 +02:00
greatroar
b10acd2af7 Test and benchmark blob sorting in internal/repository 2020-07-25 12:12:58 +02:00
Alexander Weiss
9175795fdb Check contents in archiver
When backing up with a parent snapshot and the file is not changed, also
check if contents are still available in index.
2020-07-25 08:18:28 +02:00
rawtaz
5d8d70542f
Merge pull request #2852 from MichaelEischer/drop-go-1.11
Drop support for Go version 1.11
2020-07-25 01:17:36 +02:00
Michael Eischer
7c23381a2b Drop support for Go version 1.11 2020-07-24 18:52:39 +02:00
MichaelEischer
34181b13a2
Merge pull request #2328 from MichaelEischer/no-repeated-checks
Fix duplicate tree checks within `restic check`
2020-07-22 22:08:02 +02:00
Alexander Weiss
a666a6d576 Add tests and merge indexes in index benchmarks 2020-07-22 21:54:02 +02:00
Alexander Weiss
e388d962a5 Merge final indexes together for faster index access 2020-07-22 21:54:02 +02:00
Alexander Weiss
3b7a3711e6 Add more realistic index benchmarks
+ reduce test size of BenchmarkMasterIndexLookupParallel
2020-07-21 07:18:20 +02:00
Michael Eischer
9b0e718852 checker: Test that blob types are not confused 2020-07-20 23:43:47 +02:00
MichaelEischer
82c908871d
Merge pull request #2812 from greatroar/chaining
Chaining hash table for repository.Index
2020-07-20 23:29:36 +02:00
Michael Eischer
ddf0b8cd0b checker: Properly distinguish between data and tree blobs
If a data blob and a tree blob with the same ID (= same content) exist,
then the checker did not report a data or tree blob as unused when the
blob of the other type was still in use.
2020-07-20 22:58:39 +02:00
Michael Eischer
2d0c138c9b checker: Test that check only decodes trees once
The `DuplicateTree` flag is necessary to ensure that failures cannot be
swallowed. The old checker implementation ignores errors from LoadTree
if the corresponding tree was already checked.
2020-07-20 22:51:53 +02:00