changegroup: remove reordering control (BC)
This logic - including the experimental bundle.reorder option -
was originally added in
a8e3931e3fb5 in 2011 and then later ported
to changegroup.py.
The intent of this option and associated logic is to control
the ordering of revisions in deltagroups in changegroups. At the
time it was implemented, only changegroup version 1 existed
and generaldelta revlogs were just coming into the world. Changegroup
version 1 requires that deltas be made against the last revision
sent over the wire. Used with generaldelta, this created an
impedance mismatch of sorts and resulted in changegroup producers
spending a lot of time recomputing deltas.
Revision reordering was introduced so outgoing revisions would be
sent in "generaldelta order" and producers would be able to
reuse internal deltas from storage.
Later on, we introduced changegroup version 2. It supported denoting
which revision a delta was against. So we no longer needed to
sort outgoing revisions to ensure optimal delta generation from the
producer. So, subsequent changegroup versions disabled reordering.
We also later made the changelog not store deltas by default. And
we also made the changelog send out deltas in storage order. Why we
do this for changelog, I'm not sure. Maybe we want to preserve revision
order across clones? It doesn't really matter for this commit.
Fast forward to 2018. We want to abstract storage backends. And having
changegroup code require knowledge about how deltas are stored
internally interferes with that goal.
This commit removes reordering control from changegroup generation.
After this commit, the reordering behavior is:
* The changelog is always sent out in storage order (no behavior
change).
* Non-changelog generaldelta revlogs are reordered to always be in DAG
topological order (previously, generaldelta revlogs would be emitted
in storage order for version 2 and 3 changegroups).
* Non-changelog non-generaldelta revlogs are sent in storage order (no
behavior change).
* There exists no config option to override behavior.
The big difference here is that generaldelta revlogs now *always* have
their revisions sorted in DAG order before going out over the wire. This
behavior was previously only done for changegroup version 1. Version 2
and version 3 changegroups disabled reordering because the interchange
format supported encoding arbitrary delta parents, so reordering wasn't
strictly necessary.
I can think of a few significant implications for this change.
Because changegroup receivers will now see non-changelog revisions
in DAG order instead of storage order, the internal storage order of
manifests and files may differ substantially between producer and
consumer. I don't think this matters that much, since the storage
order of manifests and files is largely hidden from users. Only
the storage order of changelog matters (because `hg log` shows the
changelog in storage order). I don't think there should be any
controversy here.
The reordering of revisions has implications for changegroup producers.
Previously, generaldelta revlogs would be emitted in storage order.
And in the common case, the internally-stored delta could effectively
be copied from disk into the deltagroup delta. This meant that emitting
delta groups for generaldelta revlogs would be mostly linear read I/O.
This is desirable for performance. With us now reordering generaldelta
revlog revisions in DAG order, the read operations may use more random
I/O instead of sequential I/O. This could result in performance
loss. But with the prevalence of SSDs and fast random I/O, I'm not
too worried. (Note: the optimal emission order for revlogs is actually
delta encoding order. But the changegroup code wasn't doing that before
or after this change. We could potentially implement that in a later
commit.)
Changegroups in DAG order will have implications for receivers.
Previously, receiving storage order might mean seeing a number of
interleaved branches. This would mean long delta chains, sparse
I/O, and possibly more fulltext revisions instead of deltas, blowing
up storage storage. (This is the same set of problems that sparse
revlogs aims to address.) With the producer now sending revisions in DAG
order, the receiver also stores revisions in DAG order. That means
revisions for the same DAG branch are all grouped together. And this
should yield better storage outcomes. In other words, sending the
reordered changegroup allows the receiver to have better storage
order and for the producer to not propagate its (possibly sub-optimal)
internal storage order.
On the mozilla-unified repository, this change influences bundle
generation:
$ hg bundle -t none-v2 -a
before: time: real 355.680 secs (user 256.790+0.000 sys 16.820+0.000)
after: time: real 382.950 secs (user 281.700+0.000 sys 17.690+0.000)
before: 7,150,228,967 bytes (uncompressed)
after: 7,041,556,273 bytes (uncompressed)
before: 1,669,063,234 bytes (zstd l=3)
after: 1,628,598,830 bytes (zstd l=3)
$ hg unbundle
before: time: real 511.910 secs (user 466.750+0.000 sys 32.680+0.000)
after: time: real 487.790 secs (user 443.940+0.000 sys 30.840+0.000)
00manifest.d size:
source: 274,924,292 bytes
before: 304,741,626 bytes
after: 245,252,087 bytes
.hg/store total file size:
source: 2,649,133,490
before: 2,680,888,130
after: 2,627,875,673
We see the bundle size drop. That's probably because if a revlog
internally isn't storing a delta, it will choose to delta against
the last emitted revision. And on repos with interleaved branches
(like mozilla-unified), the previous revision could be an unrelated
branch and therefore be a large delta. But with this patch, the
previous revision is likely p1 or p2 and a delta should be small.
We also see the manifest size drop by ~50 MB. It's worth noting that
the manifest actually *increased* in size by ~25 MB in the old
strategy and decreased ~25 MB from its source in the new strategy.
Again, my explanation for this is that the DAG ordering in the
changegroup is resulting in better grouping of revisions in the
receiver, which results in more compact delta chains and higher
storage efficiency.
Unbundle time also dropped. I suspect this is due to the revlog having
to work less to compute deltas since the incoming deltas are more
optimal. i.e. the receiver spends less time resolving fulltext
revisions as incoming deltas bounce around between DAG branches and
delta chains.
We also see bundle generation time increase. This is not desirable.
However, the regression is only significant on the original repository:
if we generate a bundle from the repository created from the new,
always reordered bundles, we're close to baseline (if not at it with
expected noise):
$ hg bundle -t none-v2 -a
before (original): time: real 355.680 secs (user 256.790+0.000 sys 16.820+0.000)
after (original): time: real 382.950 secs (user 281.700+0.000 sys 17.690+0.000)
after (new repo): time: real 362.280 secs (user 260.300+0.000 sys 17.700+0.000)
This regression is a bit worrying because it will impact serving
canonical repositories (that don't have optimal internal storage
unless they are reordered - possibly as part of running
`hg debugupgraderepo`). However, this regression will only be
noticed by very large changegroups. And I'm guessing/hoping that
any repository that large is using clonebundles to mitigate server
load.
Again, sending DAG order isn't the optimal send order for servers:
sending in storage-delta order is. But in order to enable
storage-optimal send order, we'll need a storage API that handles
sorting. Future commits will introduce such an API.
Differential Revision: https://phab.mercurial-scm.org/D4721
Testing changing branch on commits
==================================
Setup
$ cat >> $HGRCPATH << EOF
> [alias]
> glog = log -G -T "{rev}:{node|short} {desc}\n{branch} ({bookmarks})"
> [experimental]
> evolution = createmarkers
> [extensions]
> rebase=
> EOF
$ hg init repo
$ cd repo
$ for ch in a b c d e; do echo foo >> $ch; hg ci -Aqm "Added "$ch; done
$ hg glog
@ 4:aa98ab95a928 Added e
| default ()
o 3:62615734edd5 Added d
| default ()
o 2:28ad74487de9 Added c
| default ()
o 1:29becc82797a Added b
| default ()
o 0:18d04c59bb5d Added a
default ()
$ hg branches
default 4:aa98ab95a928
Try without passing a new branch name
$ hg branch -r .
abort: no branch name specified for the revisions
[255]
Setting an invalid branch name
$ hg branch -r . a:b
abort: ':' cannot be used in a name
[255]
$ hg branch -r . tip
abort: the name 'tip' is reserved
[255]
$ hg branch -r . 1234
abort: cannot use an integer as a name
[255]
Change on non-linear set of commits
$ hg branch -r 2 -r 4 foo
abort: cannot change branch of non-linear revisions
[255]
Change in middle of the stack (linear commits)
$ hg branch -r 1::3 foo
abort: cannot change branch of changeset with children
[255]
Change with dirty working directory
$ echo bar > a
$ hg branch -r . foo
abort: uncommitted changes
[255]
$ hg revert --all
reverting a
Change on empty revision set
$ hg branch -r 'draft() - all()' foo
abort: empty revision set
[255]
Changing branch on linear set of commits from head
Without obsmarkers
$ hg branch -r 3:4 foo --config experimental.evolution=!
changed branch on 2 changesets
saved backup bundle to $TESTTMP/repo/.hg/strip-backup/62615734edd5-e86bd13a-branch-change.hg
$ hg glog
@ 4:3938acfb5c0f Added e
| foo ()
o 3:9435da006bdc Added d
| foo ()
o 2:28ad74487de9 Added c
| default ()
o 1:29becc82797a Added b
| default ()
o 0:18d04c59bb5d Added a
default ()
$ hg branches
foo 4:3938acfb5c0f
default 2:28ad74487de9 (inactive)
With obsmarkers
$ hg branch -r 3::4 bar
changed branch on 2 changesets
$ hg glog
@ 6:7c1991464886 Added e
| bar ()
o 5:1ea05e93925f Added d
| bar ()
o 2:28ad74487de9 Added c
| default ()
o 1:29becc82797a Added b
| default ()
o 0:18d04c59bb5d Added a
default ()
$ hg branches
bar 6:7c1991464886
default 2:28ad74487de9 (inactive)
Change branch name to an existing branch
$ hg branch -r . default
abort: a branch of the same name already exists
[255]
Changing on a branch head which is not topological head
$ hg branch -r 2 stable
abort: cannot change branch of changeset with children
[255]
Enabling the allowunstable config and trying to change branch on a branch head
which is not a topological head
$ echo "[experimental]" >> .hg/hgrc
$ echo "evolution.allowunstable=yes" >> .hg/hgrc
$ hg branch -r 2 foo
changed branch on 1 changesets
2 new orphan changesets
Changing branch of an obsoleted changeset
$ hg branch -r 4 foobar
abort: hidden revision '4' was rewritten as: 7c1991464886!
(use --hidden to access hidden revisions)
[255]
$ hg branch -r 4 --hidden foobar
abort: cannot change branch of a obsolete changeset
[255]
Make sure bookmark movement is correct
$ hg bookmark b1
$ hg glog -r '.^::'
@ 6:7c1991464886 Added e
| bar (b1)
* 5:1ea05e93925f Added d
| bar ()
~
$ hg branch -r '(.^)::' wat --debug
changing branch of '1ea05e93925f806d875a2163f9b76764be644636' from 'bar' to 'wat'
committing files:
d
committing manifest
committing changelog
new node id is 343660ccab7400da637bd6a211d07f413536d718
changing branch of '7c19914648869f5b02fc7fed31ddee9783fdd680' from 'bar' to 'wat'
committing files:
e
committing manifest
committing changelog
new node id is de1404b45a69f8cc6437d7679033ee33e9efb4ba
moving bookmarks ['b1'] from 7c19914648869f5b02fc7fed31ddee9783fdd680 to de1404b45a69f8cc6437d7679033ee33e9efb4ba
resolving manifests
branchmerge: False, force: False, partial: False
ancestor: 7c1991464886, local: 7c1991464886+, remote: de1404b45a69
starting 4 threads for background file closing (?)
changed branch on 2 changesets
updating the branch cache
invalid branchheads cache (served): tip differs
$ hg glog -r '(.^)::'
@ 9:de1404b45a69 Added e
| wat (b1)
* 8:343660ccab74 Added d
| wat ()
~
Make sure phase handling is correct
$ echo foo >> bar
$ hg ci -Aqm "added bar" --secret
1 new orphan changesets
$ hg glog -r .
@ 10:8ad1294c1660 added bar
| wat (b1)
~
$ hg branch -r . secret
changed branch on 1 changesets
$ hg phase -r .
11: secret
$ hg branches
secret 11:38a9b2d53f98
foo 7:8a4729a5e2b8
wat 9:de1404b45a69 (inactive)
default 2:28ad74487de9 (inactive)
$ hg branch
secret
Changing branch of another head, different from one on which we are
$ hg glog
@ 11:38a9b2d53f98 added bar
| secret (b1)
* 9:de1404b45a69 Added e
| wat ()
* 8:343660ccab74 Added d
| wat ()
| o 7:8a4729a5e2b8 Added c
| | foo ()
x | 2:28ad74487de9 Added c
|/ default ()
o 1:29becc82797a Added b
| default ()
o 0:18d04c59bb5d Added a
default ()
$ hg branch
secret
$ hg branch -r 7 foobar
changed branch on 1 changesets
The current branch must be preserved
$ hg branch
secret
Changing branch on multiple heads at once
$ hg rebase -s 8 -d 12 --keepbranches -q
$ hg rebase -s 14 -d 1 --keepbranches -q
$ hg branch -r 0: stable
changed branch on 6 changesets
$ hg glog
@ 23:6a5ddbcfb870 added bar
| stable (b1)
o 22:baedc6e98a67 Added e
| stable ()
| o 21:99ac7bf8aad1 Added d
| | stable ()
| o 20:0ecb4d39c4bd Added c
|/ stable ()
o 19:fd45b986b109 Added b
| stable ()
o 18:204d2769eca2 Added a
stable ()
$ hg branches
stable 23:6a5ddbcfb870
$ hg branch
stable
Changing to same branch is no-op
$ hg branch -r 19::21 stable
changed branch on 0 changesets
Changing branch name to existing branch name if the branch of parent of root of
revs is same as the new branch name
$ hg branch -r 20::21 bugfix
changed branch on 2 changesets
$ hg glog
o 25:714defe1cf34 Added d
| bugfix ()
o 24:98394def28fc Added c
| bugfix ()
| @ 23:6a5ddbcfb870 added bar
| | stable (b1)
| o 22:baedc6e98a67 Added e
|/ stable ()
o 19:fd45b986b109 Added b
| stable ()
o 18:204d2769eca2 Added a
stable ()
$ hg branch -r 24:25 stable
changed branch on 2 changesets
$ hg glog
o 27:4ec342341562 Added d
| stable ()
o 26:83f48859c2de Added c
| stable ()
| @ 23:6a5ddbcfb870 added bar
| | stable (b1)
| o 22:baedc6e98a67 Added e
|/ stable ()
o 19:fd45b986b109 Added b
| stable ()
o 18:204d2769eca2 Added a
stable ()
Testing on merge
$ hg merge -r 26
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
(branch merge, don't forget to commit)
$ hg branch -r . abcd
abort: outstanding uncommitted merge
[255]
$ hg ci -m "Merge commit"
$ hg branch -r '(.^)::' def
abort: cannot change branch of a merge commit
[255]
Changing branch on public changeset
$ hg phase -r 27 -p
$ hg branch -r 27 def
abort: cannot change branch of public changesets
(see 'hg help phases' for details)
[255]