Mercurial > hg
view tests/test-lfs-largefiles.t @ 42044:bb271ec2fbfb
compression: introduce a `storage.revlog.zstd.level` configuration
This option control the zstd compression level used when compressing revlog
chunk. The usage of zstd for revlog compression has not graduated from
experimental yet, but we intend to fix that soon.
The option name for the compression level is more straight forward to pick, so
this changesets comes first. Having a dedicated option for each compression
engine is useful because they don't support the same range of values.
I ran the same measurement as for the zlib compression level (in the parent
changesets). The variation in repository size is stay mostly in the same (small)
range. The "read/write" performance see smallish variation, but are overall much
better than zlib. Write performance show the same tend of having better write
performance for when reaching high-end compression.
Again, we don't intend to change the default zstd compression level (currently:
3) in this series. However this is worth investigating in the future.
The Performance comparison of zlib vs zstd is quite impressive. The repository
size stay in the same range, but the performance are much better in all
situations.
Comparison summary
==================
We are looking at:
- performance range for zlib
- performance range for zstd
- comparison of default zstd (level-3) to default zlib (level 6)
- comparison of the slowest zstd time to the fastest zlib time
Read performance:
-----------------
| zlib | zstd | cmp | f2s
mercurial | 0.170159 - 0.189219 | 0.144127 - 0.149624 | 80% | 88%
pypy | 2.679217 - 2.768691 | 1.532317 - 1.705044 | 60% | 63%
netbeans | 122.477027 - 141.620281 | 72.996346 - 89.731560 | 58% | 73%
mozilla | 147.867662 - 170.572118 | 91.700995 - 105.853099 | 56% | 71%
Write performance:
------------------
| zlib | zstd | cmp | f2s
mercurial | 53.250304 - 56.2936129 | 40.877025 - 45.677286 | 75% | 86%
pypy | 460.721984 - 476.589918 | 270.545409 - 301.002219 | 63% | 65%
netbeans | 520.560316 - 715.930400 | 370.356311 - 428.329652 | 55% | 82%
mozilla | 739.803002 - 987.056093 | 505.152906 - 591.930683 | 57% | 80%
Raw data
--------
repo alg lvl .hg/store size 00manifest.d read write
mercurial zlib 1 49,402,813 5,963,475 0.170159 53.250304
mercurial zlib 6 47,197,397 5,875,730 0.182820 56.264320
mercurial zlib 9 47,121,596 5,849,781 0.189219 56.293612
mercurial zstd 1 49,737,084 5,966,355 0.144127 40.877025
mercurial zstd 3 48,961,867 5,895,208 0.146376 42.268142
mercurial zstd 5 48,200,592 5,938,676 0.149624 43.162875
mercurial zstd 10 47,833,520 5,913,353 0.145185 44.012489
mercurial zstd 15 47,314,604 5,728,679 0.147686 45.677286
mercurial zstd 20 47,330,502 5,830,539 0.145789 45.025407
mercurial zstd 22 47,330,076 5,830,539 0.143996 44.690460
pypy zlib 1 370,830,572 28,462,425 2.679217 460.721984
pypy zlib 6 340,112,317 27,648,747 2.768691 467.537158
pypy zlib 9 338,360,736 27,639,003 2.763495 476.589918
pypy zstd 1 362,377,479 27,916,214 1.532317 270.545409
pypy zstd 3 354,137,693 27,905,988 1.686718 294.951509
pypy zstd 5 342,640,043 27,655,774 1.705044 301.002219
pypy zstd 10 334,224,327 27,164,493 1.567287 285.186239
pypy zstd 15 329,000,363 26,645,965 1.637729 299.561332
pypy zstd 20 324,534,039 26,199,547 1.526813 302.149827
pypy zstd 22 324,530,595 26,198,932 1.525718 307.821218
netbeans zlib 1 1,281,847,810 165,495,457 122.477027 520.560316
netbeans zlib 6 1,205,284,353 159,161,207 139.876147 715.930400
netbeans zlib 9 1,197,135,671 155,034,586 141.620281 678.297064
netbeans zstd 1 1,259,581,737 160,840,613 72.996346 370.356311
netbeans zstd 3 1,232,978,122 157,691,551 81.622317 396.733087
netbeans zstd 5 1,208,034,075 160,246,880 83.080549 364.342626
netbeans zstd 10 1,188,624,176 156,083,417 79.323935 403.594602
netbeans zstd 15 1,176,973,589 153,859,477 89.731560 428.329652
netbeans zstd 20 1,162,958,258 151,147,535 82.842667 392.335349
netbeans zstd 22 1,162,707,029 151,150,220 82.565695 402.840655
mozilla zlib 1 2,775,497,186 298,527,987 147.867662 751.263721
mozilla zlib 6 2,596,856,420 286,597,671 170.572118 987.056093
mozilla zlib 9 2,587,542,494 287,018,264 163.622338 739.803002
mozilla zstd 1 2,723,159,348 286,617,532 91.700995 570.042751
mozilla zstd 3 2,665,055,001 286,152,013 95.240155 561.412805
mozilla zstd 5 2,607,819,817 288,060,030 101.978048 505.152906
mozilla zstd 10 2,558,761,085 283,967,648 104.113481 497.771202
mozilla zstd 15 2,526,216,060 275,581,300 105.853099 591.930683
mozilla zstd 20 2,485,114,806 266,478,859 95.268795 576.515389
mozilla zstd 22 2,484,869,080 266,456,505 94.429282 572.785537
author | Pierre-Yves David <pierre-yves.david@octobus.net> |
---|---|
date | Wed, 27 Mar 2019 18:35:59 +0100 |
parents | a0886a4d6dce |
children | 84a93fa7ecfd |
line wrap: on
line source
#require no-reposimplestore no-chg This tests the interaction between the largefiles and lfs extensions, and conversion from largefiles -> lfs. $ cat >> $HGRCPATH << EOF > [extensions] > largefiles = > > [lfs] > # standin files are 41 bytes. Stay bigger for clarity. > threshold = 42 > EOF Setup a repo with a normal file and a largefile, above and below the lfs threshold to test lfconvert. *.txt start life as a normal file; *.bin start as an lfs/largefile. $ hg init largefiles $ cd largefiles $ echo 'normal' > normal.txt $ echo 'normal above lfs threshold 0000000000000000000000000' > lfs.txt $ hg ci -Am 'normal.txt' adding lfs.txt adding normal.txt $ echo 'largefile' > large.bin $ echo 'largefile above lfs threshold 0000000000000000000000' > lfs.bin $ hg add --large large.bin lfs.bin $ hg ci -m 'add largefiles' $ cat >> $HGRCPATH << EOF > [extensions] > lfs = > EOF Add an lfs file and normal file that collide with files on the other branch. large.bin is added as a normal file, and is named as such only to clash with the largefile on the other branch. $ hg up -q '.^' $ echo 'below lfs threshold' > large.bin $ echo 'lfs above the lfs threshold for length 0000000000000' > lfs.bin $ hg ci -Am 'add with lfs extension' adding large.bin adding lfs.bin created new head $ hg log -G @ changeset: 2:e989d0fa3764 | tag: tip | parent: 0:29361292f54d | user: test | date: Thu Jan 01 00:00:00 1970 +0000 | summary: add with lfs extension | | o changeset: 1:6513aaab9ca0 |/ user: test | date: Thu Jan 01 00:00:00 1970 +0000 | summary: add largefiles | o changeset: 0:29361292f54d user: test date: Thu Jan 01 00:00:00 1970 +0000 summary: normal.txt -------------------------------------------------------------------------------- Merge largefiles into lfs branch The largefiles extension will prompt to use the normal or largefile when merged into the lfs files. `hg manifest` will show standins if present. They aren't, because largefiles merge doesn't merge content. If it did, selecting (n)ormal would convert to lfs on commit, if appropriate. BUG: Largefiles isn't running the merge tool, like when two lfs files are merged. This is probably by design, but it should probably at least prompt if content should be taken from (l)ocal or (o)ther as well. $ hg --config ui.interactive=True merge 6513aaab9ca0 <<EOF > n > n > EOF remote turned local normal file large.bin into a largefile use (l)argefile or keep (n)ormal file? n remote turned local normal file lfs.bin into a largefile use (l)argefile or keep (n)ormal file? n 0 files updated, 0 files merged, 0 files removed, 0 files unresolved (branch merge, don't forget to commit) $ hg ci -m 'merge lfs with largefiles -> normal' $ hg manifest large.bin lfs.bin lfs.txt normal.txt The merged lfs.bin resolved to lfs because the (n)ormal option was picked. The lfs.txt file is unchanged by the merge, because it was added before lfs was enabled, and the content didn't change. $ hg debugdata lfs.bin 0 version https://git-lfs.github.com/spec/v1 oid sha256:81c7492b2c05e130431f65a87651b54a30c5da72c99ce35a1e9b9872a807312b size 53 x-is-binary 0 $ hg debugdata lfs.txt 0 normal above lfs threshold 0000000000000000000000000 Another filelog entry is NOT made by the merge, so nothing is committed as lfs. $ hg log -r . -T '{join(lfs_files, ", ")}\n' Replay the last merge, but pick (l)arge this time. The manifest will show any standins. $ hg up -Cq e989d0fa3764 $ hg --config ui.interactive=True merge 6513aaab9ca0 <<EOF > l > l > EOF remote turned local normal file large.bin into a largefile use (l)argefile or keep (n)ormal file? l remote turned local normal file lfs.bin into a largefile use (l)argefile or keep (n)ormal file? l getting changed largefiles 2 largefiles updated, 0 removed 2 files updated, 0 files merged, 2 files removed, 0 files unresolved (branch merge, don't forget to commit) $ hg ci -m 'merge lfs with largefiles -> large' created new head $ hg manifest .hglf/large.bin .hglf/lfs.bin lfs.txt normal.txt -------------------------------------------------------------------------------- Merge lfs into largefiles branch $ hg up -Cq 6513aaab9ca0 $ hg --config ui.interactive=True merge e989d0fa3764 <<EOF > n > n > EOF remote turned local largefile large.bin into a normal file keep (l)argefile or use (n)ormal file? n remote turned local largefile lfs.bin into a normal file keep (l)argefile or use (n)ormal file? n getting changed largefiles 0 largefiles updated, 0 removed 2 files updated, 0 files merged, 2 files removed, 0 files unresolved (branch merge, don't forget to commit) $ hg ci -m 'merge largefiles with lfs -> normal' created new head $ hg manifest large.bin lfs.bin lfs.txt normal.txt The merged lfs.bin got converted to lfs because the (n)ormal option was picked. The lfs.txt file is unchanged by the merge, because it was added before lfs was enabled. $ hg debugdata lfs.bin 0 version https://git-lfs.github.com/spec/v1 oid sha256:81c7492b2c05e130431f65a87651b54a30c5da72c99ce35a1e9b9872a807312b size 53 x-is-binary 0 $ hg debugdata lfs.txt 0 normal above lfs threshold 0000000000000000000000000 Another filelog entry is NOT made by the merge, so nothing is committed as lfs. $ hg log -r . -T '{join(lfs_files, ", ")}\n' Replay the last merge, but pick (l)arge this time. The manifest will show the standins. $ hg up -Cq 6513aaab9ca0 $ hg --config ui.interactive=True merge e989d0fa3764 <<EOF > l > l > EOF remote turned local largefile large.bin into a normal file keep (l)argefile or use (n)ormal file? l remote turned local largefile lfs.bin into a normal file keep (l)argefile or use (n)ormal file? l 0 files updated, 0 files merged, 0 files removed, 0 files unresolved (branch merge, don't forget to commit) $ hg ci -m 'merge largefiles with lfs -> large' created new head $ hg manifest .hglf/large.bin .hglf/lfs.bin lfs.txt normal.txt -------------------------------------------------------------------------------- When both largefiles and lfs are configured to add by size, the tie goes to largefiles since it hooks cmdutil.add() and lfs hooks the filelog write in the commit. By the time the commit occurs, the tracked file is smaller than the threshold (assuming it is > 41, so the standins don't become lfs objects). $ "$PYTHON" -c 'import sys ; sys.stdout.write("y\n" * 1048576)' > large_by_size.bin $ hg --config largefiles.minsize=1 ci -Am 'large by size' adding large_by_size.bin as a largefile $ hg manifest .hglf/large.bin .hglf/large_by_size.bin .hglf/lfs.bin lfs.txt normal.txt $ hg rm large_by_size.bin $ hg ci -m 'remove large_by_size.bin' Largefiles doesn't do anything special with diff, so it falls back to diffing the standins. Extdiff also is standin based comparison. Diff and extdiff both work on the original file for lfs objects. Largefile -> lfs transition $ hg diff -r 1 -r 3 diff -r 6513aaab9ca0 -r dcc5ce63e252 .hglf/large.bin --- a/.hglf/large.bin Thu Jan 01 00:00:00 1970 +0000 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,1 +0,0 @@ -cef9a458373df9b0743a0d3c14d0c66fb19b8629 diff -r 6513aaab9ca0 -r dcc5ce63e252 .hglf/lfs.bin --- a/.hglf/lfs.bin Thu Jan 01 00:00:00 1970 +0000 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,1 +0,0 @@ -557fb6309cef935e1ac2c8296508379e4b15a6e6 diff -r 6513aaab9ca0 -r dcc5ce63e252 large.bin --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/large.bin Thu Jan 01 00:00:00 1970 +0000 @@ -0,0 +1,1 @@ +below lfs threshold diff -r 6513aaab9ca0 -r dcc5ce63e252 lfs.bin --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/lfs.bin Thu Jan 01 00:00:00 1970 +0000 @@ -0,0 +1,1 @@ +lfs above the lfs threshold for length 0000000000000 lfs -> largefiles transition $ hg diff -r 2 -r 6 diff -r e989d0fa3764 -r 95e1e80325c8 .hglf/large.bin --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/.hglf/large.bin Thu Jan 01 00:00:00 1970 +0000 @@ -0,0 +1,1 @@ +cef9a458373df9b0743a0d3c14d0c66fb19b8629 diff -r e989d0fa3764 -r 95e1e80325c8 .hglf/lfs.bin --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/.hglf/lfs.bin Thu Jan 01 00:00:00 1970 +0000 @@ -0,0 +1,1 @@ +557fb6309cef935e1ac2c8296508379e4b15a6e6 diff -r e989d0fa3764 -r 95e1e80325c8 large.bin --- a/large.bin Thu Jan 01 00:00:00 1970 +0000 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,1 +0,0 @@ -below lfs threshold diff -r e989d0fa3764 -r 95e1e80325c8 lfs.bin --- a/lfs.bin Thu Jan 01 00:00:00 1970 +0000 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,1 +0,0 @@ -lfs above the lfs threshold for length 0000000000000 A largefiles repo can be converted to lfs. The lfconvert command uses the convert extension under the hood with --to-normal. So the --config based parameters are available, but not --authormap, --branchmap, etc. $ cd .. $ hg lfconvert --to-normal largefiles nolargefiles 2>&1 initializing destination nolargefiles 0 additional largefiles cached scanning source... sorting... converting... 8 normal.txt 7 add largefiles 6 add with lfs extension 5 merge lfs with largefiles -> normal 4 merge lfs with largefiles -> large 3 merge largefiles with lfs -> normal 2 merge largefiles with lfs -> large 1 large by size 0 remove large_by_size.bin $ cd nolargefiles The requirement is added to the destination repo. $ cat .hg/requires dotencode fncache generaldelta lfs revlogv1 sparserevlog store $ hg log -r 'all()' -G -T '{rev} {join(lfs_files, ", ")} ({desc})\n' o 8 large_by_size.bin (remove large_by_size.bin) | o 7 large_by_size.bin (large by size) | o 6 (merge largefiles with lfs -> large) |\ +---o 5 (merge largefiles with lfs -> normal) | |/ +---o 4 lfs.bin (merge lfs with largefiles -> large) | |/ +---o 3 (merge lfs with largefiles -> normal) | |/ | o 2 lfs.bin (add with lfs extension) | | o | 1 lfs.bin (add largefiles) |/ o 0 lfs.txt (normal.txt) $ hg debugdata lfs.bin 0 version https://git-lfs.github.com/spec/v1 oid sha256:2172a5bd492dd41ec533b9bb695f7691b6351719407ac797f0ccad5348c81e62 size 53 x-is-binary 0 $ hg debugdata lfs.bin 1 version https://git-lfs.github.com/spec/v1 oid sha256:81c7492b2c05e130431f65a87651b54a30c5da72c99ce35a1e9b9872a807312b size 53 x-is-binary 0 $ hg debugdata lfs.bin 2 version https://git-lfs.github.com/spec/v1 oid sha256:2172a5bd492dd41ec533b9bb695f7691b6351719407ac797f0ccad5348c81e62 size 53 x-is-binary 0 $ hg debugdata lfs.bin 3 abort: invalid revision identifier 3 [255] No diffs when comparing merge and p1 that kept p1's changes. Diff of lfs to largefiles no longer operates in standin files. This `head -n 20` looks dumb (since we expect no output), but if something breaks you can get 1048576 lines of +y in the output, which takes a looooooong time to print. $ hg diff -r 2:3 | head -n 20 $ hg diff -r 2:6 diff -r e989d0fa3764 -r 752e3a0d8488 large.bin --- a/large.bin Thu Jan 01 00:00:00 1970 +0000 +++ b/large.bin Thu Jan 01 00:00:00 1970 +0000 @@ -1,1 +1,1 @@ -below lfs threshold +largefile diff -r e989d0fa3764 -r 752e3a0d8488 lfs.bin --- a/lfs.bin Thu Jan 01 00:00:00 1970 +0000 +++ b/lfs.bin Thu Jan 01 00:00:00 1970 +0000 @@ -1,1 +1,1 @@ -lfs above the lfs threshold for length 0000000000000 +largefile above lfs threshold 0000000000000000000000