Mercurial > hg
view tests/test-rebase-parameters.t @ 40326:fed697fa1734
sqlitestore: file storage backend using SQLite
This commit provides an extension which uses SQLite to store file
data (as opposed to revlogs).
As the inline documentation describes, there are still several
aspects to the extension that are incomplete. But it's a start.
The extension does support basic clone, checkout, and commit
workflows, which makes it suitable for simple use cases.
One notable missing feature is support for "bundlerepos." This is
probably responsible for the most test failures when the extension
is activated as part of the test suite.
All revision data is stored in SQLite. Data is stored as zstd
compressed chunks (default if zstd is available), zlib compressed
chunks (default if zstd is not available), or raw chunks (if
configured or if a compressed delta is not smaller than the raw
delta). This makes things very similar to revlogs.
Unlike revlogs, the extension doesn't yet enforce a limit on delta
chain length. This is an obvious limitation and should be addressed.
This is somewhat mitigated by the use of zstd, which is much faster
than zlib to decompress.
There is a dedicated table for storing deltas. Deltas are stored
by the SHA-1 hash of their uncompressed content. The "fileindex" table
has columns that reference the delta for each revision and the base
delta that delta should be applied against. A recursive SQL query
is used to resolve the delta chain along with the delta data.
By storing deltas by hash, we are able to de-duplicate delta storage!
With revlogs, the same deltas in different revlogs would result in
duplicate storage of that delta. In this scheme, inserting the
duplicate delta is a no-op and delta chains simply reference the
existing delta.
When initially implementing this extension, I did not have
content-indexed deltas and deltas could be duplicated across files
(just like revlogs). When I implemented content-indexed deltas, the
size of the SQLite database for a full clone of mozilla-unified
dropped:
before: 2,554,261,504 bytes
after: 2,488,754,176 bytes
Surprisingly, this is still larger than the bytes size of revlog
files:
revlog files: 2,104,861,230 bytes
du -b: 2,254,381,614
I would have expected storage to be smaller since we're not limiting
delta chain length and since we're using zstd instead of zlib. I
suspect the SQLite indexes and per-column overhead account for the
bulk of the differences. (Keep in mind that revlog uses a 64-byte
packed struct for revision index data and deltas are stored without
padding. Aside from the 12 unused bytes in the 32 byte node field,
revlogs are pretty efficient.) Another source of overhead is file
name storage. With revlogs, file names are stored in the filesystem.
But with SQLite, we need to store file names in the database. This is
roughly equivalent to the size of the fncache file, which for the
mozilla-unified repository is ~34MB.
Since the SQLite database isn't append-only and since delta chains
can reference any delta, this opens some interesting possibilities.
For example, we could store deltas in reverse, such that fulltexts
are stored for newer revisions and deltas are applied to reconstruct
older revisions. This is likely a more optimal storage strategy for
version control, as new data tends to be more frequently accessed
than old data. We would obviously need wire protocol support for
transferring revision data from newest to oldest. And we would
probably need some kind of mechanism for "re-encoding" stores. But
it should be doable.
This extension is very much experimental quality. There are a handful
of features that don't work. It probably isn't suitable for day-to-day
use. But it could be used in limited cases (e.g. read-only checkouts
like in CI). And it is also a good proving ground for alternate
storage backends. As we continue to define interfaces for all things
storage, it will be useful to have a viable alternate storage backend
to see how things shake out in practice.
test-storage.py passes on Python 2 and introduces no new test failures on
Python 3. Having the storage-level unit tests has proved to be insanely
useful when developing this extension. Those tests caught numerous bugs
during development and I'm convinced this style of testing is the way
forward for ensuring alternate storage backends work as intended. Of
course, test coverage isn't close to what it needs to be. But it is
a start. And what coverage we have gives me confidence that basic store
functionality is implemented properly.
Differential Revision: https://phab.mercurial-scm.org/D4928
author | Gregory Szorc <gregory.szorc@gmail.com> |
---|---|
date | Tue, 09 Oct 2018 08:50:13 -0700 |
parents | 89630d0b3e23 |
children | 09680349cc2d |
line wrap: on
line source
$ cat >> $HGRCPATH <<EOF > [extensions] > rebase= > > [phases] > publish=False > > [alias] > tglog = log -G --template "{rev}: {node|short} '{desc}' {branches}\n" > EOF $ hg init a $ cd a $ hg unbundle "$TESTDIR/bundles/rebase.hg" adding changesets adding manifests adding file changes added 8 changesets with 7 changes to 7 files (+2 heads) new changesets cd010b8cd998:02de42196ebe (8 drafts) (run 'hg heads' to see heads, 'hg merge' to merge) $ hg up tip 3 files updated, 0 files merged, 0 files removed, 0 files unresolved $ echo I > I $ hg ci -AmI adding I $ hg tglog @ 8: e7ec4e813ba6 'I' | o 7: 02de42196ebe 'H' | | o 6: eea13746799a 'G' |/| o | 5: 24b6387c8c8c 'F' | | | o 4: 9520eea781bc 'E' |/ | o 3: 32af7686d403 'D' | | | o 2: 5fddd98957c8 'C' | | | o 1: 42ccdea3bb16 'B' |/ o 0: cd010b8cd998 'A' $ cd .. Version with only two heads (to allow default destination to work) $ hg clone -q -u . a a2heads -r 3 -r 8 These fail: $ hg clone -q -u . a a0 $ cd a0 $ hg rebase -s 8 -d 7 nothing to rebase [1] $ hg rebase --continue --abort abort: cannot use --abort with --continue [255] $ hg rebase --continue --collapse abort: cannot use collapse with continue or abort [255] $ hg rebase --continue --dest 4 abort: abort and continue do not allow specifying revisions [255] $ hg rebase --base 5 --source 4 abort: cannot specify both a source and a base [255] $ hg rebase --rev 5 --source 4 abort: cannot specify both a revision and a source [255] $ hg rebase --base 5 --rev 4 abort: cannot specify both a revision and a base [255] $ hg rebase --base 6 abort: branch 'default' has 3 heads - please rebase to an explicit rev (run 'hg heads .' to see heads) [255] $ hg rebase --rev '1 & !1' --dest 8 empty "rev" revision set - nothing to rebase [1] $ hg rebase --source '1 & !1' --dest 8 empty "source" revision set - nothing to rebase [1] $ hg rebase --base '1 & !1' --dest 8 empty "base" revision set - can't compute rebase set [1] $ hg rebase --dest 8 nothing to rebase - working directory parent is also destination [1] $ hg rebase -b . --dest 8 nothing to rebase - e7ec4e813ba6 is both "base" and destination [1] $ hg up -q 7 $ hg rebase --dest 8 --traceback nothing to rebase - working directory parent is already an ancestor of destination e7ec4e813ba6 [1] $ hg rebase --dest 8 -b. nothing to rebase - "base" 02de42196ebe is already an ancestor of destination e7ec4e813ba6 [1] $ hg rebase --dest '1 & !1' abort: empty revision set [255] These work: Rebase with no arguments (from 3 onto 8): $ cd .. $ hg clone -q -u . a2heads a1 $ cd a1 $ hg up -q -C 3 $ hg rebase rebasing 1:42ccdea3bb16 "B" rebasing 2:5fddd98957c8 "C" rebasing 3:32af7686d403 "D" saved backup bundle to $TESTTMP/a1/.hg/strip-backup/42ccdea3bb16-3cb021d3-rebase.hg $ hg tglog @ 6: ed65089c18f8 'D' | o 5: 7621bf1a2f17 'C' | o 4: 9430a62369c6 'B' | o 3: e7ec4e813ba6 'I' | o 2: 02de42196ebe 'H' | o 1: 24b6387c8c8c 'F' | o 0: cd010b8cd998 'A' Try to rollback after a rebase (fail): $ hg rollback no rollback information available [1] $ cd .. Rebase with base == '.' => same as no arguments (from 3 onto 8): $ hg clone -q -u 3 a2heads a2 $ cd a2 $ hg rebase --base . rebasing 1:42ccdea3bb16 "B" rebasing 2:5fddd98957c8 "C" rebasing 3:32af7686d403 "D" saved backup bundle to $TESTTMP/a2/.hg/strip-backup/42ccdea3bb16-3cb021d3-rebase.hg $ hg tglog @ 6: ed65089c18f8 'D' | o 5: 7621bf1a2f17 'C' | o 4: 9430a62369c6 'B' | o 3: e7ec4e813ba6 'I' | o 2: 02de42196ebe 'H' | o 1: 24b6387c8c8c 'F' | o 0: cd010b8cd998 'A' $ cd .. Rebase with dest == branch(.) => same as no arguments (from 3 onto 8): $ hg clone -q -u 3 a a3 $ cd a3 $ hg rebase --dest 'branch(.)' rebasing 1:42ccdea3bb16 "B" rebasing 2:5fddd98957c8 "C" rebasing 3:32af7686d403 "D" saved backup bundle to $TESTTMP/a3/.hg/strip-backup/42ccdea3bb16-3cb021d3-rebase.hg $ hg tglog @ 8: ed65089c18f8 'D' | o 7: 7621bf1a2f17 'C' | o 6: 9430a62369c6 'B' | o 5: e7ec4e813ba6 'I' | o 4: 02de42196ebe 'H' | | o 3: eea13746799a 'G' |/| o | 2: 24b6387c8c8c 'F' | | | o 1: 9520eea781bc 'E' |/ o 0: cd010b8cd998 'A' $ cd .. Specify only source (from 2 onto 8): $ hg clone -q -u . a2heads a4 $ cd a4 $ hg rebase --source 'desc("C")' rebasing 2:5fddd98957c8 "C" rebasing 3:32af7686d403 "D" saved backup bundle to $TESTTMP/a4/.hg/strip-backup/5fddd98957c8-f9244fa1-rebase.hg $ hg tglog o 6: 7726e9fd58f7 'D' | o 5: 72c8333623d0 'C' | @ 4: e7ec4e813ba6 'I' | o 3: 02de42196ebe 'H' | o 2: 24b6387c8c8c 'F' | | o 1: 42ccdea3bb16 'B' |/ o 0: cd010b8cd998 'A' $ cd .. Specify only dest (from 3 onto 6): $ hg clone -q -u 3 a a5 $ cd a5 $ hg rebase --dest 6 rebasing 1:42ccdea3bb16 "B" rebasing 2:5fddd98957c8 "C" rebasing 3:32af7686d403 "D" saved backup bundle to $TESTTMP/a5/.hg/strip-backup/42ccdea3bb16-3cb021d3-rebase.hg $ hg tglog @ 8: 8eeb3c33ad33 'D' | o 7: 2327fea05063 'C' | o 6: e4e5be0395b2 'B' | | o 5: e7ec4e813ba6 'I' | | | o 4: 02de42196ebe 'H' | | o | 3: eea13746799a 'G' |\| | o 2: 24b6387c8c8c 'F' | | o | 1: 9520eea781bc 'E' |/ o 0: cd010b8cd998 'A' $ cd .. Specify only base (from 1 onto 8): $ hg clone -q -u . a2heads a6 $ cd a6 $ hg rebase --base 'desc("D")' rebasing 1:42ccdea3bb16 "B" rebasing 2:5fddd98957c8 "C" rebasing 3:32af7686d403 "D" saved backup bundle to $TESTTMP/a6/.hg/strip-backup/42ccdea3bb16-3cb021d3-rebase.hg $ hg tglog o 6: ed65089c18f8 'D' | o 5: 7621bf1a2f17 'C' | o 4: 9430a62369c6 'B' | @ 3: e7ec4e813ba6 'I' | o 2: 02de42196ebe 'H' | o 1: 24b6387c8c8c 'F' | o 0: cd010b8cd998 'A' $ cd .. Specify source and dest (from 2 onto 7): $ hg clone -q -u . a a7 $ cd a7 $ hg rebase --source 2 --dest 7 rebasing 2:5fddd98957c8 "C" rebasing 3:32af7686d403 "D" saved backup bundle to $TESTTMP/a7/.hg/strip-backup/5fddd98957c8-f9244fa1-rebase.hg $ hg tglog o 8: 668acadedd30 'D' | o 7: 09eb682ba906 'C' | | @ 6: e7ec4e813ba6 'I' |/ o 5: 02de42196ebe 'H' | | o 4: eea13746799a 'G' |/| o | 3: 24b6387c8c8c 'F' | | | o 2: 9520eea781bc 'E' |/ | o 1: 42ccdea3bb16 'B' |/ o 0: cd010b8cd998 'A' $ cd .. Specify base and dest (from 1 onto 7): $ hg clone -q -u . a a8 $ cd a8 $ hg rebase --base 3 --dest 7 rebasing 1:42ccdea3bb16 "B" rebasing 2:5fddd98957c8 "C" rebasing 3:32af7686d403 "D" saved backup bundle to $TESTTMP/a8/.hg/strip-backup/42ccdea3bb16-3cb021d3-rebase.hg $ hg tglog o 8: 287cc92ba5a4 'D' | o 7: 6824f610a250 'C' | o 6: 7c6027df6a99 'B' | | @ 5: e7ec4e813ba6 'I' |/ o 4: 02de42196ebe 'H' | | o 3: eea13746799a 'G' |/| o | 2: 24b6387c8c8c 'F' | | | o 1: 9520eea781bc 'E' |/ o 0: cd010b8cd998 'A' $ cd .. Specify only revs (from 2 onto 8) $ hg clone -q -u . a2heads a9 $ cd a9 $ hg rebase --rev 'desc("C")::' rebasing 2:5fddd98957c8 "C" rebasing 3:32af7686d403 "D" saved backup bundle to $TESTTMP/a9/.hg/strip-backup/5fddd98957c8-f9244fa1-rebase.hg $ hg tglog o 6: 7726e9fd58f7 'D' | o 5: 72c8333623d0 'C' | @ 4: e7ec4e813ba6 'I' | o 3: 02de42196ebe 'H' | o 2: 24b6387c8c8c 'F' | | o 1: 42ccdea3bb16 'B' |/ o 0: cd010b8cd998 'A' $ cd .. Rebasing both a single revision and a merge in one command $ hg clone -q -u . a aX $ cd aX $ hg rebase -r 3 -r 6 --dest 8 rebasing 3:32af7686d403 "D" rebasing 6:eea13746799a "G" saved backup bundle to $TESTTMP/aX/.hg/strip-backup/eea13746799a-ad273fd6-rebase.hg $ cd .. Test --tool parameter: $ hg init b $ cd b $ echo c1 > c1 $ hg ci -Am c1 adding c1 $ echo c2 > c2 $ hg ci -Am c2 adding c2 $ hg up -q 0 $ echo c2b > c2 $ hg ci -Am c2b adding c2 created new head $ cd .. $ hg clone -q -u . b b1 $ cd b1 $ hg rebase -s 2 -d 1 --tool internal:local rebasing 2:e4e3f3546619 "c2b" (tip) note: rebase of 2:e4e3f3546619 created no changes to commit saved backup bundle to $TESTTMP/b1/.hg/strip-backup/e4e3f3546619-b0841178-rebase.hg $ hg cat c2 c2 $ cd .. $ hg clone -q -u . b b2 $ cd b2 $ hg rebase -s 2 -d 1 --tool internal:other rebasing 2:e4e3f3546619 "c2b" (tip) saved backup bundle to $TESTTMP/b2/.hg/strip-backup/e4e3f3546619-b0841178-rebase.hg $ hg cat c2 c2b $ cd .. $ hg clone -q -u . b b3 $ cd b3 $ hg rebase -s 2 -d 1 --tool internal:fail rebasing 2:e4e3f3546619 "c2b" (tip) unresolved conflicts (see hg resolve, then hg rebase --continue) [1] $ hg summary parent: 1:56daeba07f4b c2 parent: 2:e4e3f3546619 tip c2b branch: default commit: 1 modified, 1 unresolved (merge) update: (current) phases: 3 draft rebase: 0 rebased, 1 remaining (rebase --continue) $ hg resolve -l U c2 $ hg resolve -m c2 (no more unresolved files) continue: hg rebase --continue $ hg graft --continue abort: no graft in progress (continue: hg rebase --continue) [255] $ hg rebase -c --tool internal:fail rebasing 2:e4e3f3546619 "c2b" (tip) note: rebase of 2:e4e3f3546619 created no changes to commit saved backup bundle to $TESTTMP/b3/.hg/strip-backup/e4e3f3546619-b0841178-rebase.hg $ hg rebase -i abort: interactive history editing is supported by the 'histedit' extension (see "hg --config extensions.histedit= help -e histedit") [255] $ hg rebase --interactive abort: interactive history editing is supported by the 'histedit' extension (see "hg --config extensions.histedit= help -e histedit") [255] $ cd .. No common ancestor $ hg init separaterepo $ cd separaterepo $ touch a $ hg commit -Aqm a $ hg up -q null $ touch b $ hg commit -Aqm b $ hg rebase -d 0 nothing to rebase from d7486e00c6f1 to 3903775176ed [1] $ cd ..