Some codepaths in "git pack-objects" were not ready to use an
existing pack bitmap; now they are and as the result they have
become faster.
* ks/pack-objects-bitmap:
pack-objects: use reachability bitmap index when generating non-stdout pack
pack-objects: respect --local/--honor-pack-keep/--incremental when bitmap is in use
Starting from 6b8fda2d (pack-objects: use bitmaps when packing objects)
if a repository has bitmap index, pack-objects can nicely speedup
"Counting objects" graph traversal phase. That however was done only for
case when resultant pack is sent to stdout, not written into a file.
The reason here is for on-disk repack by default we want:
- to produce good pack (with bitmap index not-yet-packed objects are
emitted to pack in suboptimal order).
- to use more robust pack-generation codepath (avoiding possible
bugs in bitmap code and possible bitmap index corruption).
Jeff King further explains:
The reason for this split is that pack-objects tries to determine how
"careful" it should be based on whether we are packing to disk or to
stdout. Packing to disk implies "git repack", and that we will likely
delete the old packs after finishing. We want to be more careful (so
as not to carry forward a corruption, and to generate a more optimal
pack), and we presumably run less frequently and can afford extra CPU.
Whereas packing to stdout implies serving a remote via "git fetch" or
"git push". This happens more frequently (e.g., a server handling many
fetching clients), and we assume the receiving end takes more
responsibility for verifying the data.
But this isn't always the case. One might want to generate on-disk
packfiles for a specialized object transfer. Just using "--stdout" and
writing to a file is not optimal, as it will not generate the matching
pack index.
So it would be useful to have some way of overriding this heuristic:
to tell pack-objects that even though it should generate on-disk
files, it is still OK to use the reachability bitmaps to do the
traversal.
So we can teach pack-objects to use bitmap index for initial object
counting phase when generating resultant pack file too:
- if we take care to not let it be activated under git-repack:
See above about repack robustness and not forward-carrying corruption.
- if we know bitmap index generation is not enabled for resultant pack:
The current code has singleton bitmap_git, so it cannot work
simultaneously with two bitmap indices.
We also want to avoid (at least with current implementation)
generating bitmaps off of bitmaps. The reason here is: when generating
a pack, not-yet-packed objects will be emitted into pack in
suboptimal order and added to tail of the bitmap as "extended entries".
When the resultant pack + some new objects in associated repository
are in turn used to generate another pack with bitmap, the situation
repeats: new objects are again not emitted optimally and just added to
bitmap tail - not in recency order.
So the pack badness can grow over time when at each step we have
bitmapped pack + some other objects. That's why we want to avoid
generating bitmaps off of bitmaps, not to let pack badness grow.
- if we keep pack reuse enabled still only for "send-to-stdout" case:
Because pack-to-file needs to generate index for destination pack, and
currently on pack reuse raw entries are directly written out to the
destination pack by write_reused_pack(), bypassing needed for pack index
generation bookkeeping done by regular codepath in write_one() and
friends.
( In the future we might teach pack-reuse code about cases when index
also needs to be generated for resultant pack and remove
pack-reuse-only-for-stdout limitation )
This way for pack-objects -> file we get nice speedup:
erp5.git[1] (~230MB) extracted from ~ 5GB lab.nexedi.com backup
repository managed by git-backup[2] via
time echo 0186ac99 | git pack-objects --revs erp5pack
before: 37.2s
after: 26.2s
And for `git repack -adb` packed git.git
time echo 5c589a73 | git pack-objects --revs gitpack
before: 7.1s
after: 3.6s
i.e. it can be 30% - 50% speedup for pack extraction.
git-backup extracts many packs on repositories restoration. That was my
initial motivation for the patch.
[1] https://lab.nexedi.com/nexedi/erp5
[2] https://lab.nexedi.com/kirr/git-backup
NOTE
Jeff also suggests that pack.useBitmaps was probably a mistake to
introduce originally. This way we are not adding another config point,
but instead just always default to-file pack-objects not to use bitmap
index: Tools which need to generate on-disk packs with using bitmap, can
pass --use-bitmap-index explicitly. And git-repack does never pass
--use-bitmap-index, so this way we can be sure regular on-disk repacking
remains robust.
NOTE2
`git pack-objects --stdout >file.pack` + `git index-pack file.pack` is much slower
than `git pack-objects file.pack`. Extracting erp5.git pack from
lab.nexedi.com backup repository:
$ time echo 0186ac99 | git pack-objects --stdout --revs >erp5pack-stdout.pack
real 0m22.309s
user 0m21.148s
sys 0m0.932s
$ time git index-pack erp5pack-stdout.pack
real 0m50.873s <-- more than 2 times slower than time to generate pack itself!
user 0m49.300s
sys 0m1.360s
So the time for
`pack-object --stdout >file.pack` + `index-pack file.pack` is 72s,
while
`pack-objects file.pack` which does both pack and index is 27s.
And even
`pack-objects --no-use-bitmap-index file.pack` is 37s.
Jeff explains:
The packfile does not carry the sha1 of the objects. A receiving
index-pack has to compute them itself, including inflating and applying
all of the deltas.
that's why for `git-backup restore` we want to teach `git pack-objects
file.pack` to use bitmaps instead of using `git pack-objects --stdout
>file.pack` + `git index-pack file.pack`.
NOTE3
The speedup is now tracked via t/perf/p5310-pack-bitmaps.sh
Test 56dfeb62 this tree
--------------------------------------------------------------------------------
5310.2: repack to disk 8.98(8.05+0.29) 9.05(8.08+0.33) +0.8%
5310.3: simulated clone 2.02(2.27+0.09) 2.01(2.25+0.08) -0.5%
5310.4: simulated fetch 0.81(1.07+0.02) 0.81(1.05+0.04) +0.0%
5310.5: pack to file 7.58(7.04+0.28) 7.60(7.04+0.30) +0.3%
5310.6: pack to file (bitmap) 7.55(7.02+0.28) 3.25(2.82+0.18) -57.0%
5310.8: clone (partial bitmap) 1.83(2.26+0.12) 1.82(2.22+0.14) -0.5%
5310.9: pack to file (partial bitmap) 6.86(6.58+0.30) 2.87(2.74+0.20) -58.2%
More context:
http://marc.info/?t=146792101400001&r=1&w=2http://public-inbox.org/git/20160707190917.20011-1-kirr@nexedi.com/T/#t
Cc: Vicent Marti <tanoku@gmail.com>
Helped-by: Jeff King <peff@peff.net>
Signed-off-by: Kirill Smelkov <kirr@nexedi.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since 6b8fda2d (pack-objects: use bitmaps when packing objects) there
are two codepaths in pack-objects: with & without using bitmap
reachability index.
However add_object_entry_from_bitmap(), despite its non-bitmapped
counterpart add_object_entry(), in no way does check for whether --local
or --honor-pack-keep or --incremental should be respected. In
non-bitmapped codepath this is handled in want_object_in_pack(), but
bitmapped codepath has simply no such checking at all.
The bitmapped codepath however was allowing to pass in all those options
and with bitmap indices still being used under such conditions -
potentially giving wrong output (e.g. including objects from non-local or
.keep'ed pack).
We can easily fix this by noting the following: when an object comes to
add_object_entry_from_bitmap() it can come for two reasons:
1. entries coming from main pack covered by bitmap index, and
2. object coming from, possibly alternate, loose or other packs.
"2" can be already handled by want_object_in_pack() and to cover
"1" we can teach want_object_in_pack() to expect that *found_pack can be
non-NULL, meaning calling client already found object's pack entry.
In want_object_in_pack() we care to start the checks from already found
pack, if we have one, this way determining the answer right away
in case neither --local nor --honour-pack-keep are active. In
particular, as p5310-pack-bitmaps.sh shows (3 consecutive runs), we do
not do harm to served-with-bitmap clones performance-wise:
Test 56dfeb62 this tree
-----------------------------------------------------------------
5310.2: repack to disk 9.08(8.20+0.25) 9.09(8.14+0.32) +0.1%
5310.3: simulated clone 1.92(2.12+0.08) 1.93(2.12+0.09) +0.5%
5310.4: simulated fetch 0.82(1.07+0.04) 0.82(1.06+0.04) +0.0%
5310.6: partial bitmap 1.96(2.42+0.13) 1.95(2.40+0.15) -0.5%
Test 56dfeb62 this tree
-----------------------------------------------------------------
5310.2: repack to disk 9.11(8.16+0.32) 9.11(8.19+0.28) +0.0%
5310.3: simulated clone 1.93(2.14+0.07) 1.92(2.11+0.10) -0.5%
5310.4: simulated fetch 0.82(1.06+0.04) 0.82(1.04+0.05) +0.0%
5310.6: partial bitmap 1.95(2.38+0.16) 1.94(2.39+0.14) -0.5%
Test 56dfeb62 this tree
-----------------------------------------------------------------
5310.2: repack to disk 9.13(8.17+0.31) 9.07(8.13+0.28) -0.7%
5310.3: simulated clone 1.92(2.13+0.07) 1.91(2.12+0.06) -0.5%
5310.4: simulated fetch 0.82(1.08+0.03) 0.82(1.08+0.03) +0.0%
5310.6: partial bitmap 1.96(2.43+0.14) 1.96(2.42+0.14) +0.0%
with delta timings showing they are all within noise from run to run.
In the general case we do not want to call find_pack_entry_one() more than
once, because it is expensive. This patch splits the loop in
want_object_in_pack() into two parts: finding the object and seeing if it
impacts our choice to include it in the pack. We may call the inexpensive
want_found_object() twice, but we will never call find_pack_entry_one() if we
do not need to.
I appreciate help and discussing this change with Junio C Hamano and
Jeff King.
Signed-off-by: Kirill Smelkov <kirr@nexedi.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This enables JGIT to be used as a prereq in invocations of
test_expect_success (and other functions) in other test scripts.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If you ask rev-list for:
git rev-list --count --use-bitmap-index HEAD
we optimize out the actual traversal and just give you the
number of bits set in the commit bitmap. This is faster,
which is good.
But if you ask to limit the size of the traversal, like:
git rev-list --count --use-bitmap-index -n 100 HEAD
we'll still output the full bitmapped number we found. On
the surface, that might even seem OK. You explicitly asked
to use the bitmap index, and it was cheap to compute the
real answer, so we gave it to you.
But there's something much more complicated going on under
the hood. If we don't have a bitmap directly for HEAD, then
we have to actually traverse backwards, looking for a
bitmapped commit. And _that_ traversal is bounded by our
`-n` count.
This is a good thing, because it bounds the work we have to
do, which is probably what the user wanted by asking for
`-n`. But now it makes the output quite confusing. You might
get many values:
- your `-n` value, if we walked back and never found a
bitmap (or fewer if there weren't that many commits)
- the actual full count, if we found a bitmap root for
every path of our traversal with in the `-n` limit
- any number in between! We might have walked back and
found _some_ bitmaps, but then cut off the traversal
early with some commits not accounted for in the result.
So you cannot even see a value higher than your `-n` and say
"OK, bitmaps kicked in, this must be the real full count".
The only sane thing is for git to just clamp the value to a
maximum of the `-n` value, which means we should output the
exact same results whether bitmaps are in use or not.
The test in t5310 demonstrates this by using `-n 1`.
Without this patch we fail in the full-bitmap case (where we
do not have to traverse at all) but _not_ in the
partial-bitmap case (where we have to walk down to find an
actual bitmap). With this patch, both cases just work.
I didn't implement the crazy in-between case, just because
it's complicated to set up, and is really a subset of the
full-count case, which we do cover.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A minor bugfix when pack bitmap is used with "rev-list --count".
* jk/rev-list-no-bitmap-while-pruning:
rev-list: disable --use-bitmap-index when pruning commits
The reachability bitmaps do not have enough information to
tell us which commits might have changed path "foo", so the
current code produces wrong answers for:
git rev-list --use-bitmap-index --count HEAD -- foo
(it silently ignores the "foo" limiter). Instead, we should
fall back to doing a normal traversal (it is OK to fall
back rather than complain, because --use-bitmap-index is a
pure optimization, and might not kick in for other reasons,
such as there being no bitmaps in the repository).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Splitting pack-objects output into multiple packs is incompatible
with the use of reachability bitmap.
* jk/pack-objects-no-bitmap-when-splitting:
pack-objects: turn off bitmaps when we split packs
If a pack.packSizeLimit is set, we may split the pack data
across multiple packfiles. This means we cannot generate
.bitmap files, as they require that all of the reachable
objects are in the same pack. We check that condition when
we are generating the list of objects to pack (and disable
bitmaps if we are not packing everything), but we forgot to
update it when we notice that we needed to split (which
doesn't happen until the actual write phase).
The resulting bitmaps are quite bogus (they mention entries
that do not exist in the pack!) and can cause a fetch or
push to send insufficient objects.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We currently have pack.writeBitmaps, which originally
operated at the pack-objects level. This should really have
been a repack.* option from day one. Let's give it the more
sensible name, but keep the old version as a deprecated
synonym.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When pack-objects is computing the reachability bitmap to
serve a fetch request, it can erroneously die() if some of
the UNINTERESTING objects are not present. Upload-pack
throws away HAVE lines from the client for objects we do not
have, but we may have a tip object without all of its
ancestors (e.g., if the tip is no longer reachable and was
new enough to survive a `git prune`, but some of its
reachable objects did get pruned).
In the non-bitmap case, we do a revision walk with the HAVE
objects marked as UNINTERESTING. The revision walker
explicitly ignores errors in accessing UNINTERESTING commits
to handle this case (and we do not bother looking at
UNINTERESTING trees or blobs at all).
When we have bitmaps, however, the process is quite
different. The bitmap index for a pack-objects run is
calculated in two separate steps:
First, we perform an extensive walk from all the HAVEs to
find the full set of objects reachable from them. This walk
is usually optimized away because we are expected to hit an
object with a bitmap during the traversal, which allows us
to terminate early.
Secondly, we perform an extensive walk from all the WANTs,
which usually also terminates early because we hit a commit
with an existing bitmap.
Once we have the resulting bitmaps from the two walks, we
AND-NOT them together to obtain the resulting set of objects
we need to pack.
When we are walking the HAVE objects, the revision walker
does not know that we are walking it only to mark the
results as uninteresting. We strip out the UNINTERESTING flag,
because those objects _are_ interesting to us during the
first walk. We want to keep going to get a complete set of
reachable objects if we can.
We need some way to tell the revision walker that it's OK to
silently truncate the HAVE walk, just like it does for the
UNINTERESTING case. This patch introduces a new
`ignore_missing_links` flag to the `rev_info` struct, which
we set only for the HAVE walk.
It also adds tests to cover UNINTERESTING objects missing
from several positions: a missing blob, a missing tree, and
a missing parent commit. The missing blob already worked (as
we do not care about its contents at all), but the other two
cases caused us to die().
Note that there are a few cases we do not need to test:
1. We do not need to test a missing tree, with the blob
still present. Without the tree that refers to it, we
would not know that the blob is relevant to our walk.
2. We do not need to test a tip commit that is missing.
Upload-pack omits these for us (and in fact, we
complain even in the non-bitmap case if it fails to do
so).
Reported-by: Siddharth Agarwal <sid0@fb.com>
Signed-off-by: Vicent Marti <tanoku@gmail.com>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The pack bitmap format requires that we have a single bit
for each object in the pack, and that each object's bitmap
represents its complete set of reachable objects. Therefore
we have no way to represent the bitmap of an object which
references objects outside the pack.
We notice this problem while generating the bitmaps, as we
try to find the offset of a particular object and realize
that we do not have it. In this case we die, and neither the
bitmap nor the pack is generated. This is correct, but
perhaps a little unfriendly. If you have bitmaps turned on
in the config, many repacks will fail which would otherwise
succeed. E.g., incremental repacks, repacks with "-l" when
you have alternates, ".keep" files.
Instead, this patch notices early that we are omitting some
objects from the pack and turns off bitmaps (with a
warning). Note that this is not strictly correct, as it's
possible that the object being omitted is not reachable from
any other object in the pack. In practice, this is almost
never the case, and there are two advantages to doing it
this way:
1. The code is much simpler, as we do not have to cleanly
abort the bitmap-generation process midway through.
2. We do not waste time partially generating bitmaps only
to find out that some object deep in the history is not
being packed.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we use pack bitmaps rather than walking the object
graph, we end up with the list of objects to include in the
packfile, but we do not know the path at which any tree or
blob objects would be found.
In a recently packed repository, this is fine. A fetch would
use the paths only as a heuristic in the delta compression
phase, and a fully packed repository should not need to do
much delta compression.
As time passes, though, we may acquire more objects on top
of our large bitmapped pack. If clients fetch frequently,
then they never even look at the bitmapped history, and all
works as usual. However, a client who has not fetched since
the last bitmap repack will have "have" tips in the
bitmapped history, but "want" newer objects.
The bitmaps themselves degrade gracefully in this
circumstance. We manually walk the more recent bits of
history, and then use bitmaps when we hit them.
But we would also like to perform delta compression between
the newer objects and the bitmapped objects (both to delta
against what we know the user already has, but also between
"new" and "old" objects that the user is fetching). The lack
of pathnames makes our delta heuristics much less effective.
This patch adds an optional cache of the 32-bit name_hash
values to the end of the bitmap file. If present, a reader
can use it to match bitmapped and non-bitmapped names during
delta compression.
Here are perf results for p5310:
Test origin/master HEAD^ HEAD
-------------------------------------------------------------------------------------------------
5310.2: repack to disk 36.81(37.82+1.43) 47.70(48.74+1.41) +29.6% 47.75(48.70+1.51) +29.7%
5310.3: simulated clone 30.78(29.70+2.14) 1.08(0.97+0.10) -96.5% 1.07(0.94+0.12) -96.5%
5310.4: simulated fetch 3.16(6.10+0.08) 3.54(10.65+0.06) +12.0% 1.70(3.07+0.06) -46.2%
5310.6: partial bitmap 36.76(43.19+1.81) 6.71(11.25+0.76) -81.7% 4.08(6.26+0.46) -88.9%
You can see that the time spent on an incremental fetch goes
down, as our delta heuristics are able to do their work.
And we save time on the partial bitmap clone for the same
reason.
Signed-off-by: Vicent Marti <tanoku@gmail.com>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Now that we can read and write bitmaps, we can exercise them
with some basic functionality tests. These tests aren't
particularly useful for seeing the benefit, as the test
repo is too small for it to make a difference. However, we
can at least check that using bitmaps does not break anything.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
