Commit Graph

1244 Commits

Author SHA1 Message Date
Alexander Neumann 1dd4b9b60e
Merge pull request #3788 from greatroar/sftp-posix-rename
backend/sftp: Support atomic rename
2022-06-06 19:39:48 +02:00
Alexander Neumann 07114ccb21
Merge pull request #3789 from greatroar/fix-loadblob
internal/repository: Fix LoadBlob + fuzz test
2022-06-06 19:33:42 +02:00
greatroar c9557b2822 internal/repository: Fix LoadBlob + fuzz test
When given a buf that is big enough for a compressed blob but not its
decompressed contents, the copy at the end of LoadBlob would skip the
last part of the contents.

Fixes #3783.
2022-06-06 17:02:28 +02:00
greatroar fa8f02292e backend/sftp: Support atomic rename
... if the server has posix-rename@openssh.com.
OpenSSH introduced this extension in 2008:
7c29661471
2022-06-06 13:40:42 +02:00
Michael Eischer 5eba1217e7 migrate: Cleanup option to request repository check 2022-06-04 23:45:00 +02:00
MichaelEischer 60ca6b1418
Merge pull request #3774 from greatroar/archiver-pool
archiver: Remove cleanup goroutine from BufferPool
2022-06-04 18:50:24 +02:00
greatroar 0db1d11b2e archiver: Remove cleanup goroutine from BufferPool
This isn't doing anything. Channels should get cleaned up by the GC when
the last reference to them disappears, just like all other data
structures. Also inlined BufferPool.Put in Buffer.Release, its only
caller.
2022-05-29 17:09:16 +02:00
Alexander Neumann d2c5843c68
Merge pull request #3704 from MichaelEischer/compression-migrations
Support migration to repository format with compression
2022-05-29 15:52:21 +02:00
Alexander Neumann 78a21bbccf
Merge pull request #3752 from MichaelEischer/fix-dir-sync-errors
local: Ignore additional errors for directory syncing
2022-05-29 12:54:51 +02:00
greatroar dde8e9e296 internal/restic: Custom ID.MarshalJSON
This skips an allocation. internal/archiver benchmarks, Linux/amd64:

name                     old time/op    new time/op    delta
ArchiverSaveFileSmall-8    3.94ms ± 6%    3.91ms ± 6%    ~     (p=0.947 n=20+20)
ArchiverSaveFileLarge-8     304ms ± 3%     301ms ± 4%    ~     (p=0.265 n=18+18)

name                     old speed      new speed      delta
ArchiverSaveFileSmall-8  1.04MB/s ± 6%  1.05MB/s ± 6%    ~     (p=0.803 n=20+20)
ArchiverSaveFileLarge-8   142MB/s ± 3%   143MB/s ± 4%    ~     (p=0.421 n=18+19)

name                     old alloc/op   new alloc/op   delta
ArchiverSaveFileSmall-8    17.9MB ± 0%    17.9MB ± 0%  -0.01%  (p=0.000 n=19+19)
ArchiverSaveFileLarge-8     382MB ± 2%     382MB ± 1%    ~     (p=0.687 n=20+19)

name                     old allocs/op  new allocs/op  delta
ArchiverSaveFileSmall-8       540 ± 1%       528 ± 0%  -2.19%  (p=0.000 n=19+19)
ArchiverSaveFileLarge-8     1.93k ± 3%     1.79k ± 4%  -7.06%  (p=0.000 n=20+20)
2022-05-27 12:26:37 +02:00
MichaelEischer 88a8701fb5
Merge pull request #3734 from lbausch/validate-patterns
Validate exclude patterns
2022-05-14 16:20:15 +02:00
MichaelEischer b2a2e5f727
Merge pull request #3753 from greatroar/indexmap-alloc
repository: Re-tune indexmap allocation strategy
2022-05-14 15:44:08 +02:00
MichaelEischer b52c631bd3
Merge pull request #3754 from greatroar/simplify-hashing
hashing: Fix up comments
2022-05-14 15:33:51 +02:00
Lorenz Bausch 36bd464e8c
Add tests for validating exclude patterns 2022-05-11 22:41:00 +02:00
greatroar 39a335e690 hashing: Fix up comments 2022-05-11 21:36:10 +02:00
greatroar 5141228e0c repository: Re-tune indexmap allocation strategy
fd05037e1a changed the allocation batch
size from 256 to 128 under the assumption that an indexEntry is 60 bytes
on amd64, but it's 64: structs are padded out to a multiple of 8 for
alignment reasons. That means we'd waste no space in malloc even without
the batch allocation, at least on 64-bit machines. While that strategy
cuts the overallocation down dramatically for many small indexes, it also
seems to slow allocation down (Go 1.18, Linux, amd64, -benchtime=2s):

    name                   old time/op    new time/op    delta
    DecodeIndex-8             4.67s ± 5%     4.60s ± 1%      ~     (p=0.953 n=10+5)
    DecodeIndexParallel-8     4.67s ± 3%     4.60s ± 1%      ~     (p=0.953 n=10+5)
    IndexHasUnknown-8        37.8ns ± 8%    36.5ns ±14%      ~     (p=0.841 n=5+5)
    IndexHasKnown-8          38.5ns ±12%    37.7ns ±10%      ~     (p=0.968 n=5+5)
    IndexAlloc-8              615ms ±18%     607ms ± 1%      ~     (p=1.000 n=10+5)
    IndexAllocParallel-8      245ms ±11%     285ms ± 6%   +16.40%  (p=0.001 n=10+5)
    MasterIndexAlloc-8        286ms ± 9%     275ms ± 2%      ~     (p=1.000 n=10+5)
    LoadIndex/v1-8           27.0ms ± 4%    26.8ms ± 1%      ~     (p=0.690 n=5+5)
    LoadIndex/v2-8           22.4ms ± 1%    22.8ms ± 2%    +1.48%  (p=0.016 n=5+5)

    name                   old alloc/op   new alloc/op   delta
    IndexAlloc-8              446MB ± 0%     446MB ± 0%    -0.00%  (p=0.000 n=8+4)
    IndexAllocParallel-8      446MB ± 0%     446MB ± 0%    -0.00%  (p=0.008 n=8+5)
    MasterIndexAlloc-8        213MB ± 0%     159MB ± 0%   -25.47%  (p=0.000 n=10+5)

    name                   old allocs/op  new allocs/op  delta
    IndexAlloc-8               913k ± 0%     2632k ± 0%  +188.19%  (p=0.008 n=5+5)
    IndexAllocParallel-8       913k ± 0%     2632k ± 0%  +188.21%  (p=0.008 n=5+5)
    MasterIndexAlloc-8         318k ± 0%     1172k ± 0%  +267.86%  (p=0.008 n=5+5)

Instead, this patch sets a batch size of 4, which means no space is
wasted by malloc on 64-bit and very little on 32-bit. It still gets very
close to the savings from not allocating in batches, without requiring
special code for bits.UintSize==64. Benchmark results, again for
Linux/amd64:

    name                   old time/op    new time/op    delta
    DecodeIndex-8             4.67s ± 5%     4.83s ± 9%     ~     (p=0.315 n=10+10)
    DecodeIndexParallel-8     4.67s ± 3%     4.68s ± 4%     ~     (p=0.315 n=10+10)
    IndexHasUnknown-8        37.8ns ± 8%    44.5ns ±19%     ~     (p=0.095 n=5+5)
    IndexHasKnown-8          38.5ns ±12%    36.9ns ± 8%     ~     (p=0.690 n=5+5)
    IndexAlloc-8              615ms ±18%     628ms ±18%     ~     (p=0.218 n=10+10)
    IndexAllocParallel-8      245ms ±11%     262ms ± 9%   +7.02%  (p=0.043 n=10+10)
    MasterIndexAlloc-8        286ms ± 9%     287ms ±13%     ~     (p=1.000 n=10+10)
    LoadIndex/v1-8           27.0ms ± 4%    26.8ms ± 0%     ~     (p=1.000 n=5+5)
    LoadIndex/v2-8           22.4ms ± 1%    22.5ms ± 0%     ~     (p=0.056 n=5+5)

    name                   old alloc/op   new alloc/op   delta
    IndexAlloc-8              446MB ± 0%     446MB ± 0%     ~     (p=1.000 n=8+10)
    IndexAllocParallel-8      446MB ± 0%     446MB ± 0%   -0.00%  (p=0.000 n=8+8)
    MasterIndexAlloc-8        213MB ± 0%     160MB ± 0%  -25.02%  (p=0.000 n=10+9)

    name                   old allocs/op  new allocs/op  delta
    IndexAlloc-8               913k ± 0%     1333k ± 0%  +45.94%  (p=0.000 n=8+10)
    IndexAllocParallel-8       913k ± 0%     1333k ± 0%  +45.94%  (p=0.000 n=8+8)
    MasterIndexAlloc-8         318k ± 0%      525k ± 0%  +64.99%  (p=0.000 n=10+10)

The allocation method indexmap.newEntry has also been rewritten in a
form that is a few instructions shorter.
2022-05-11 21:22:14 +02:00
Michael Eischer 48a0d83143 local: Ignore additional errors for directory syncing
Apparently SMB/CIFS on Linux/macOS returns somewhat random errnos when
trying to sync a windows share which does not support calling fsync for
a directory.
2022-05-11 20:37:59 +02:00
MichaelEischer ac36fda155
Merge pull request #3749 from greatroar/simplify-hashing
hashing: Remove io.WriterTo implementation
2022-05-11 20:03:43 +02:00
MichaelEischer df554e5f69
Merge pull request #3748 from greatroar/runworkers
repository: Remove RunWorkers, report ctx.Err()
2022-05-11 19:38:46 +02:00
greatroar 54b8337813 hashing: Remove io.WriterTo implementation
This functionality has gone unused since
4b3dc415ef changed hashing.Reader's only
client to use ioutil.ReadAll on a bufio.Reader wrapping the hashing
Reader.

Reverts bcb852a8d0.
2022-05-10 23:41:18 +02:00
greatroar 2e0f1f5113 repository: Remove RunWorkers, report ctx.Err()
This removes RunWorkers, which had become mere overhead by successive
refactors. It also ensures that each former user of that function
returns any context error that occurs, so failure to complete an
operation is always reported as an error.
2022-05-10 22:26:00 +02:00
MichaelEischer 47c56dea5c
Merge pull request #3746 from greatroar/cache-lstat
cache: Don't Lstat before creating CACHEDIR.TAG
2022-05-10 20:31:15 +02:00
greatroar 2da377c582 cache: Don't Lstat before creating the tag file
The tag file is opened with O_CREATE|O_EXCL and ErrExist is handled, so
we don't need to check for existence first.
2022-05-10 18:52:39 +02:00
Michael Eischer ae7e51382a Fix error on temp file deletion on windows
Apparently it can take a moment between closing a tempfile marked as
DELETE_ON_CLOSE and it actually being deleted. During that time the file
is inaccessible. Thus just skip deleting the temp file on windows.
2022-05-09 22:43:26 +02:00
Michael Eischer c1bbbcd0dc migrate: Allow migrations to request a check run
This is currently only used by upgrade_repo_v2.
2022-05-09 22:31:30 +02:00
Michael Eischer 5815f727ee checker: convert error type to use pointer-receivers 2022-05-09 22:31:30 +02:00
Michael Eischer e36a40db10 upgrade_repo_v2: Use atomic replace for supported backends 2022-05-09 22:31:30 +02:00
Michael Eischer 5406743102 prune: Automatically repack uncompressed trees for repo v2
Tree packs are cached locally at clients and thus benefit a lot from
being compressed. Ensure this be having prune always repack pack files
containing uncompressed trees.
2022-05-09 22:31:30 +02:00
Alexander Neumann c8c0d659ec Add migration to compress all data 2022-05-09 22:31:30 +02:00
Alexander Neumann 8c244214bf Add tests for upgrade migration 2022-05-09 22:31:30 +02:00
Alexander Neumann a5f1d318ac Try to make repo upgrade migration more failsafe 2022-05-09 22:31:30 +02:00
Alexander Neumann 82ed5a3a15 Add repo upgrade migration 2022-05-09 22:31:30 +02:00
Michael Eischer 92816fa966 init: Enable compression support by default 2022-05-09 22:31:30 +02:00
Lorenz Bausch 9fb81c4246
Validate exclude patterns 2022-05-07 21:12:47 +02:00
Lorenz Bausch e7fd200237
Keep original pattern for later use 2022-05-07 21:08:09 +02:00
Michael Eischer cf5cb673fb repository: Use existing method to collect pack ids 2022-04-30 19:14:21 +02:00
Michael Eischer b335cb6285 repository: Refactor index IDs collection 2022-04-30 19:14:21 +02:00
Daniel Gröber f31b4f29c1 Use config file modes to derive new dir/file modes
Fixes #2351
2022-04-30 15:59:51 +02:00
Michael Eischer 4b01b06f2f repository: Test compressed blobs in StreamPack 2022-04-30 11:34:10 +02:00
Michael Eischer bcab548617 pack: slightly expand testing of compressed blobs 2022-04-30 11:34:10 +02:00
Michael Eischer ec2b25565a repository: test uncompressedLength field and index example 2022-04-30 11:34:10 +02:00
Michael Eischer 9ffb8920f1 repository: run blackbox tests using old and new repo version 2022-04-30 11:34:10 +02:00
Michael Eischer abe5935693 repository: unify repository version-specific initialization
Mark the master index as compressed also when initializing a new
repository. This is only relevant for testing.
2022-04-30 11:34:10 +02:00
Alexander Neumann 8776031f96 Leave allocating slices to the decompress code 2022-04-30 11:34:10 +02:00
Alexander Neumann 5eb05a0afe Configure zstd encoder/decoder 2022-04-30 11:34:10 +02:00
Michael Eischer 2f36e044db Cleanup pack header check 2022-04-30 11:34:10 +02:00
Alexander Neumann 8b11b86383 Add option global --compression 2022-04-30 11:34:10 +02:00
Michael Eischer 7132df529e repository: Increase index size for repo version 2
A compressed index is only about one third the size of an uncompressed
one. Thus increase the number of entries in an index to avoid cluttering
the repository with small indexes.
2022-04-30 11:34:10 +02:00
Michael Eischer 66f9048bce repository: Alloc zstd encoder/decoder on demand 2022-04-30 11:34:10 +02:00
Michael Eischer fd05037e1a repository: recalibrate index batch allocation size 2022-04-30 11:34:10 +02:00