2
2
mirror of https://github.com/octoleo/restic.git synced 2024-11-30 00:33:57 +00:00
Commit Graph

2954 Commits

Author SHA1 Message Date
Alexander Neumann
b46774be21 Merge pull request #1214 from ricardoseriani/fix-manual.rst-key-remove
Fix manual.rst to use key remove instead of key rm
2017-09-07 21:16:46 +02:00
Tobias Klein
1073bfba37 flexible grouping option for the forget-command 2017-09-06 20:14:18 +02:00
Alexander Neumann
5dfb4d1195 Merge pull request #1209 from restic/handle-colliding-names
Resolve name collisions
2017-09-05 22:33:32 +02:00
Alexander Neumann
0a2219c5f7 Travis: Reduce workload by removing Go tip 2017-09-05 22:08:23 +02:00
Alexander Neumann
ff3149831e Merge pull request #1210 from prattmic/glob_test
filter: document recursive wildcards
2017-09-05 21:50:27 +02:00
Alexander Neumann
c935d0558c Add entry to CHANGELOG 2017-09-05 21:48:13 +02:00
Alexander Neumann
83eb075e3a Resolve name collisions
At the moment when two items to be saved have the same directory name,
restic only saves the first one to the repo. Let's say we have a
structure like this:

    dir1
    └── subdir
        └── file
    dir2
    └── subdir
        └── file

When restic is run on `dir1/subdir` and `dir2/subdir`, it will only save
the first `subdir`:

    $ restic backup dir1/subdir dir2/subdir
    [...]

    $ restic ls -l latest
    drwxr-xr-x  1000   100      0 2017-08-27 20:56:39 /subdir
    -rw-r--r--  1000   100     17 2017-08-27 20:56:39 /subdir/file

That's obviously a bad thing, caused by an early decision to strip the
full path to the files/dirs to save and only leave the last directory.

This commit partly resolves this by handling colliding names and
resolving the conflicts. Restic will now append a counter to the file
(`-123`) until the conflict is resolved. So in the example above, we'll
end up with the following structure:

    $ restic ls -l latest
    drwxr-xr-x  1000   100      0 2017-08-27 20:56:39 /subdir
    -rw-r--r--  1000   100     17 2017-08-27 20:56:39 /subdir/file
    drwxr-xr-x  1000   100      0 2017-08-27 20:56:46 /subdir-1
    -rw-r--r--  1000   100     17 2017-08-27 20:56:46 /subdir-1/file

This partly addresses #549 and closes #1179.

At first I thought that the obvious correction would be to archive the
full path. But it turns out that collisions may still occur: Suppose you
have a file named `foo` in the current directory, and the parent directory
also contains a file `foo`. Archiving these with restic also causes a
collision, since restic strips the `../` from the first file:

    $ restic backup ../foo foo

This also happens with `tar`, which does not handle the collision and
will happily archive two files called `foo`.

So, the best way forward is to handle name collisions and archive the
whole path. The latter will be tackled in a separate PR.
2017-09-05 21:47:02 +02:00
Ricardo Seriani
204c2bf09c Fix manual to use key remove instead of key rm 2017-09-05 16:22:42 -03:00
Alexander Neumann
2444522243 Add test for colliding names 2017-09-05 21:10:02 +02:00
Michael Pratt
92eb1cbffd filter: document recursive wildcards
Match/ChildMatch accept a ** pattern which is not noted in the doc
string, nor do any of the docs or tests specify whether the match is
greedy (i.e., can 'foo/**/bar' match paths with additional intermediate
bar directories?).

Add a note to the doc string and add test cases for greedy matches.
2017-09-04 14:38:48 -07:00
Alexander Neumann
8c40ae5a03 Add entry to CHANGELOG 2017-09-04 21:58:33 +02:00
Alexander Neumann
fa2ee78a5c Merge pull request #1044 from lloeki/982-improve-restore
Improve restore
2017-09-04 21:51:12 +02:00
Tobias Klein
e4a5cdc5bc forget: group-by-tags-only 2017-09-03 17:11:25 +02:00
Tobias Klein
2d73a273af saving a variable 2017-09-03 17:09:55 +02:00
Tobias Klein
761af08889 fix: bucker
bucker "always" does not return a unique id in case of exact same timestamps
2017-09-03 17:09:55 +02:00
Alexander Neumann
0ee1650f82 Merge pull request #1191 from prattmic/profile
debug: properly handle interrupted profiles
2017-09-03 09:49:46 +02:00
Alexander Neumann
0e647417f3 Add entry to CHANGELOG 2017-09-03 09:49:37 +02:00
Alexander Neumann
d1bf5a4882 Merge pull request #1203 from myfreeweb/master
Handle SIGINFO on all supported platforms
2017-09-03 09:44:24 +02:00
Alexander Neumann
b8414b240c Add entry to CHANGELOG 2017-09-03 09:44:15 +02:00
Greg V
3fbdd12b04 Handle SIGINFO on all supported platforms
Not just darwin
2017-09-02 22:06:31 +03:00
Alexander Neumann
a3f6bf3e5a Merge pull request #1202 from restic/fix-manpages
Fix manpages, update Go version
2017-09-02 10:11:43 +02:00
Alexander Neumann
3a5805db50 Update Go versions for CI 2017-09-02 09:29:02 +02:00
Alexander Neumann
de8c64e767 Use deterministic date for man pages 2017-09-02 09:27:11 +02:00
Alexander Neumann
73d6b15095 Merge pull request #1201 from gjmf/patch-1
Fixed word-o. ("Package", not "packet".)
2017-09-01 21:21:18 +02:00
Alexander Neumann
5d396b9302 Merge pull request #1200 from molivier/patch-1
Update manual.rst
2017-09-01 21:21:16 +02:00
Graham Freeman
61d2519111 Fixed word-o. ("Package", not "packet".)
Fixed a word-o. homebrew is a package manager, not a packet manager. :)
2017-08-31 13:34:53 -07:00
Matthieu OLIVIER
e61c94a846 Update manual.rst
`key remove` becomes `key rm`.
2017-08-31 18:26:04 +02:00
Alexander Neumann
7ed0f61f3f Merge pull request #1189 from FiloSottile/patch-2
doc/design: fix keys.data MAC format description
2017-08-29 20:44:31 +02:00
Alexander Neumann
85055d1c68 Merge pull request #1187 from FiloSottile/patch-1
internal/crypto: small simplifications
2017-08-29 20:43:15 +02:00
Michael Pratt
e4c469c149 debug: properly handle interrupted profiles
By default (i.e., without profile.NoShutdownHook), profile.Start listens
for SIGINT and will stop the profile and call os.Exit(0).

restic already listens for SIGINT and runs its own cleanup handlers
before calling os.Exit(0).

As is, these handlers are racing when an interrupt occurs, and in my
experience, restic tends to win the race, resulting in an unusable
profile.

Eliminate the race and properly stop profiles on interrupt by disabling
package profile's signal handler and instead stop the profile in a
restic cleanup handler.
2017-08-28 22:03:26 -07:00
Filippo Valsorda
9940e8d9f1 internal/crypto: small simplifications
* append operates on len, not cap (not a bug since len is set to cap above, but let's avoid the confusion)
* no need to extend ciphertext again to cap after we made it big enough
* make consistent use of ciphertext[:ivSize] vs iv[:]
* make all input problems errors and impossible/catastrophic cases panics
2017-08-29 00:30:06 +02:00
Filippo Valsorda
3dccca1f27 doc/design: fix keys.data MAC format description
"not including the last 32 byte" was wrong, should have been 16 bytes. But the whole description is redundant anyway.
2017-08-29 00:22:11 +02:00
Alexander Neumann
22e96a37f8 Merge pull request #1184 from prattmic/docs
Doc cleanup and mention restore include/exclude
2017-08-28 21:14:48 +02:00
Alexander Neumann
48b1ab5aaf Merge pull request #1182 from restic/fix-1167
local: do not create dirs below data/ for non-existing dir
2017-08-28 21:13:24 +02:00
Alexander Neumann
0230fa188f Add entry to CHANGELOG 2017-08-28 21:13:14 +02:00
Alexander Neumann
4118ce876e Merge pull request #1185 from prattmic/gcs_panic
gs: fix nil dereference
2017-08-28 21:11:30 +02:00
Michael Pratt
9537bc561d gs: fix nil dereference
info can be nil if err != nil, resulting in a nil dereference while
logging:

$ # GCS config
$ ./restic init
debug enabled
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0x935947]

goroutine 1 [running]:
github.com/restic/restic/internal/backend/gs.(*Backend).Save(0xc420012690, 0xe84e80, 0xc420010448, 0xb57149, 0x3, 0xc4203fc140, 0x40, 0xe7be40, 0xc4201d8f90, 0xa0, ...)
	src/github.com/restic/restic/internal/backend/gs/gs.go:226 +0x6d7
github.com/restic/restic/internal/repository.AddKey(0xe84e80, 0xc420010448, 0xc4202f0360, 0xc42000a1b0, 0x4, 0x0, 0xa55b60, 0xc4203043e0, 0xa55420)
	src/github.com/restic/restic/internal/repository/key.go:235 +0x4a1
github.com/restic/restic/internal/repository.createMasterKey(0xc4202f0360, 0xc42000a1b0, 0x4, 0xa55420, 0xc420304370, 0x6a6070)
	src/github.com/restic/restic/internal/repository/key.go:62 +0x60
github.com/restic/restic/internal/repository.(*Repository).init(0xc4202f0360, 0xe84e80, 0xc420010448, 0xc42000a1b0, 0x4, 0x1, 0xc42030a440, 0x40, 0x32a4573d3d9eb5, 0x0, ...)
	src/github.com/restic/restic/internal/repository/repository.go:403 +0x5d
github.com/restic/restic/internal/repository.(*Repository).Init(0xc4202f0360, 0xe84e80, 0xc420010448, 0xc42000a1b0, 0x4, 0xe84e40, 0xc42004ad80)
	src/github.com/restic/restic/internal/repository/repository.go:397 +0x12c
main.runInit(0xc420018072, 0x16, 0x0, 0x0, 0x0, 0xe84e40, 0xc42004ad80, 0xc42000a1b0, 0x4, 0xe7dac0, ...)
	src/github.com/restic/restic/cmd/restic/cmd_init.go:47 +0x2a4
main.glob..func9(0xeb5000, 0xedad70, 0x0, 0x0, 0x0, 0x0)
	src/github.com/restic/restic/cmd/restic/cmd_init.go:20 +0x8e
github.com/restic/restic/vendor/github.com/spf13/cobra.(*Command).execute(0xeb5000, 0xedad70, 0x0, 0x0, 0xeb5000, 0xedad70)
	src/github.com/restic/restic/vendor/github.com/spf13/cobra/command.go:649 +0x457
github.com/restic/restic/vendor/github.com/spf13/cobra.(*Command).ExecuteC(0xeb3e00, 0xc420011650, 0xa55b60, 0xc420011660)
	src/github.com/restic/restic/vendor/github.com/spf13/cobra/command.go:728 +0x339
github.com/restic/restic/vendor/github.com/spf13/cobra.(*Command).Execute(0xeb3e00, 0x25, 0xc4201a7eb8)
	src/github.com/restic/restic/vendor/github.com/spf13/cobra/command.go:687 +0x2b
main.main()
	src/github.com/restic/restic/cmd/restic/main.go:72 +0x268

(The error was likely because I had just enabled the GCS API. Subsequent
runs were fine.)
2017-08-27 21:36:04 -07:00
Michael Pratt
ae43c47ca8 doc: add mention of restore --exclude/--include
There is a lot more detail that could be added here, but it is worth
getting things off the ground with at least a mention that it is
possible to restore individual files.

Updates #396
2017-08-27 18:36:00 -07:00
Michael Pratt
2fa4060991 doc: fix inconsistencies
* Replace references to ~/shared/work/web which should be ~/web.

* Replace references to ~/tmp which should be /tmp.

* Restore /home/art to /tmp/restore-art instead of /tmp/restore-work,
  which is clearly a copy/paste from the command above.
2017-08-27 18:35:01 -07:00
Alexander Neumann
f9a934759f sftp: Improve error handling for non-existing dir 2017-08-27 20:53:04 +02:00
Alexander Neumann
3686b1ffe5 local: Create directories below data/ if it exists 2017-08-27 20:52:58 +02:00
Alexander Neumann
ea017a49c3 local: Add test for #1167
It was discovered that restic creates directories when a non-existing
directory is specified as a local repository.
2017-08-27 20:38:46 +02:00
Alexander Neumann
3559f9c776 Merge pull request #1174 from pwaring/patch-1
Update minimum Go version
2017-08-25 21:26:13 +02:00
Paul Waring
637f57ca71 Update minimum Go version
Minimum version is now 1.8 according to build.go (from latest master, cloned a few minutes ago):

```
paul@voga:~/third-dev/restic$ go run build.go 
Go version go1.7.4 detected, restic requires at least Go 1.8
exit status 1
```
2017-08-24 19:51:11 +01:00
Alexander Neumann
4e60156b45 Add entry to CHANGELOG 2017-08-18 19:50:32 +02:00
Alexander Neumann
af9946b098 Merge pull request #1164 from ricardoseriani/fix-key-remove-command
Change key rm command to key remove
2017-08-18 19:49:50 +02:00
Ricardo Seriani
b7d4b0f821 Update man pages 2017-08-17 16:44:28 -03:00
Ricardo Seriani
62ed776a8c Change key rm command to key remove
Change key rm command to key remove, to follow manual and other commands
2017-08-17 11:03:26 -03:00
Loic Nageleisen
f880ff21aa Fixing restore with excluded
An exclude filter is basically a 'wildcard but foo', so even if a
childMayMatch, other children of a dir may not, therefore childMayMatch
does not matter, but we should not go down unless the dir is selected
for restore.
2017-08-16 15:25:02 +02:00
Loic Nageleisen
4a36993c19 Smarter filter when children won't match
This improves restore performance by several orders of magniture by not
going through the whole tree recursively when we can anticipate that no
match will ever occur.
2017-08-16 15:25:02 +02:00