tests/test-largefiles-wireproto.t

localrepo: experimental support for non-zlib revlog compression

The final part of integrating the compression manager APIs into
revlog storage is the plumbing for repositories to advertise they
are using non-zlib storage and for revlogs to instantiate a non-zlib
compression engine.

The main intent of the compression manager work was to zstd all
of the things. Adding zstd to revlogs has proved to be more involved
than other places because revlogs are... special. Very small inputs
and the use of delta chains (which are themselves a form of
compression) are a completely different use case from streaming
compression, which bundles and the wire protocol employ. I've
conducted numerous experiments with zstd in revlogs and have yet
to formalize compression settings and a storage architecture that
I'm confident I won't regret later. In other words, I'm not yet
ready to commit to a new mechanism for using zstd - or any other
compression format - in revlogs.

That being said, having some support for zstd (and other compression
formats) in revlogs in core is beneficial. It can allow others to
conduct experiments.

This patch introduces *highly experimental* support for non-zlib
compression formats in revlogs. Introduced is a config option to
control which compression engine to use. Also introduced is a namespace
of "exp-compression-*" requirements to denote support for non-zlib
compression in revlogs. I've prefixed the namespace with "exp-"
(short for "experimental") because I'm not confident of the
requirements "schema" and in no way want to give the illusion of
supporting these requirements in the future. I fully intend to drop
support for these requirements once we figure out what we're doing
with zstd in revlogs.

A good portion of the patch is teaching the requirements system
about registered compression engines and passing the requested
compression engine as an opener option so revlogs can instantiate
the proper compression engine for new operations.

That's a verbose way of saying "we can now use zstd in revlogs!"

On an `hg pull` conversion of the mozilla-unified repo with no extra
redelta settings (like aggressivemergedeltas), we can see the impact
of zstd vs zlib in revlogs:

$ hg perfrevlogchunks -c
! chunk
! wall 2.032052 comb 2.040000 user 1.990000 sys 0.050000 (best of 5)
! wall 1.866360 comb 1.860000 user 1.820000 sys 0.040000 (best of 6)

! chunk batch
! wall 1.877261 comb 1.870000 user 1.860000 sys 0.010000 (best of 6)
! wall 1.705410 comb 1.710000 user 1.690000 sys 0.020000 (best of 6)

$ hg perfrevlogchunks -m
! chunk
! wall 2.721427 comb 2.720000 user 2.640000 sys 0.080000 (best of 4)
! wall 2.035076 comb 2.030000 user 1.950000 sys 0.080000 (best of 5)

! chunk batch
! wall 2.614561 comb 2.620000 user 2.580000 sys 0.040000 (best of 4)
! wall 1.910252 comb 1.910000 user 1.880000 sys 0.030000 (best of 6)

$ hg perfrevlog -c -d 1
! wall 4.812885 comb 4.820000 user 4.800000 sys 0.020000 (best of 3)
! wall 4.699621 comb 4.710000 user 4.700000 sys 0.010000 (best of 3)

$ hg perfrevlog -m -d 1000
! wall 34.252800 comb 34.250000 user 33.730000 sys 0.520000 (best of 3)
! wall 24.094999 comb 24.090000 user 23.320000 sys 0.770000 (best of 3)

Only modest wins for the changelog. But manifest reading is
significantly faster. What's going on?

One reason might be data volume. zstd decompresses faster. So given
more bytes, it will put more distance between it and zlib.

Another reason is size. In the current design, zstd revlogs are
*larger*:

debugcreatestreamclonebundle (size in bytes)
zlib: 1,638,852,492
zstd: 1,680,601,332

I haven't investigated this fully, but I reckon a significant cause of
larger revlogs is that the zstd frame/header has more bytes than
zlib's. For very small inputs or data that doesn't compress well, we'll
tend to store more uncompressed chunks than with zlib (because the
compressed size isn't smaller than original). This will make revlog
reading faster because it is doing less decompression.

Moving on to bundle performance:

$ hg bundle -a -t none-v2 (total CPU time)
zlib: 102.79s
zstd:  97.75s

So, marginal CPU decrease for reading all chunks in all revlogs
(this is somewhat disappointing).

$ hg bundle -a -t <engine>-v2 (total CPU time)
zlib: 191.59s
zstd: 115.36s

This last test effectively measures the difference between zlib->zlib
and zstd->zstd for revlogs to bundle. This is a rough approximation of
what a server does during `hg clone`.

There are some promising results for zstd. But not enough for me to
feel comfortable advertising it to users. We'll get there...

This file contains testcases that tend to be related to the wire protocol part
of largefiles.

  $ USERCACHE="$TESTTMP/cache"; export USERCACHE
  $ mkdir "${USERCACHE}"
  $ cat >> $HGRCPATH <<EOF
  > [extensions]
  > largefiles=
  > purge=
  > rebase=
  > transplant=
  > [phases]
  > publish=False
  > [largefiles]
  > minsize=2
  > patterns=glob:**.dat
  > usercache=${USERCACHE}
  > [web]
  > allow_archive = zip
  > [hooks]
  > precommit=sh -c "echo \\"Invoking status precommit hook\\"; hg status"
  > EOF


#if serve
vanilla clients not locked out from largefiles servers on vanilla repos
  $ mkdir r1
  $ cd r1
  $ hg init
  $ echo c1 > f1
  $ hg add f1
  $ hg commit -m "m1"
  Invoking status precommit hook
  A f1
  $ cd ..
  $ hg serve -R r1 -d -p $HGPORT --pid-file hg.pid
  $ cat hg.pid >> $DAEMON_PIDS
  $ hg --config extensions.largefiles=! clone http://localhost:$HGPORT r2
  requesting all changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files
  updating to branch default
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved

largefiles clients still work with vanilla servers
  $ hg serve --config extensions.largefiles=! -R r1 -d -p $HGPORT1 --pid-file hg.pid
  $ cat hg.pid >> $DAEMON_PIDS
  $ hg clone http://localhost:$HGPORT1 r3
  requesting all changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files
  updating to branch default
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
#endif

vanilla clients locked out from largefiles http repos
  $ mkdir r4
  $ cd r4
  $ hg init
  $ echo c1 > f1
  $ hg add --large f1
  $ hg commit -m "m1"
  Invoking status precommit hook
  A f1
  $ cd ..

largefiles can be pushed locally (issue3583)
  $ hg init dest
  $ cd r4
  $ hg outgoing ../dest
  comparing with ../dest
  searching for changes
  changeset:   0:639881c12b4c
  tag:         tip
  user:        test
  date:        Thu Jan 01 00:00:00 1970 +0000
  summary:     m1
  
  $ hg push ../dest
  pushing to ../dest
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files

exit code with nothing outgoing (issue3611)
  $ hg outgoing ../dest
  comparing with ../dest
  searching for changes
  no changes found
  [1]
  $ cd ..

#if serve
  $ hg serve -R r4 -d -p $HGPORT2 --pid-file hg.pid
  $ cat hg.pid >> $DAEMON_PIDS
  $ hg --config extensions.largefiles=! clone http://localhost:$HGPORT2 r5
  abort: remote error:
  
  This repository uses the largefiles extension.
  
  Please enable it in your Mercurial config file.
  [255]

used all HGPORTs, kill all daemons
  $ killdaemons.py
#endif

vanilla clients locked out from largefiles ssh repos
  $ hg --config extensions.largefiles=! clone -e "python \"$TESTDIR/dummyssh\"" ssh://user@dummy/r4 r5
  remote: 
  remote: This repository uses the largefiles extension.
  remote: 
  remote: Please enable it in your Mercurial config file.
  remote: 
  remote: -
  abort: remote error
  (check previous remote output)
  [255]

#if serve

largefiles clients refuse to push largefiles repos to vanilla servers
  $ mkdir r6
  $ cd r6
  $ hg init
  $ echo c1 > f1
  $ hg add f1
  $ hg commit -m "m1"
  Invoking status precommit hook
  A f1
  $ cat >> .hg/hgrc <<!
  > [web]
  > push_ssl = false
  > allow_push = *
  > !
  $ cd ..
  $ hg clone r6 r7
  updating to branch default
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ cd r7
  $ echo c2 > f2
  $ hg add --large f2
  $ hg commit -m "m2"
  Invoking status precommit hook
  A f2
  $ hg verify --large
  checking changesets
  checking manifests
  crosschecking files in changesets and manifests
  checking files
  2 files, 2 changesets, 2 total revisions
  searching 1 changesets for largefiles
  verified existence of 1 revisions of 1 largefiles
  $ hg serve --config extensions.largefiles=! -R ../r6 -d -p $HGPORT --pid-file ../hg.pid
  $ cat ../hg.pid >> $DAEMON_PIDS
  $ hg push http://localhost:$HGPORT
  pushing to http://localhost:$HGPORT/
  searching for changes
  abort: http://localhost:$HGPORT/ does not appear to be a largefile store
  [255]
  $ cd ..

putlfile errors are shown (issue3123)
Corrupt the cached largefile in r7 and move it out of the servers usercache
  $ mv r7/.hg/largefiles/4cdac4d8b084d0b599525cf732437fb337d422a8 .
  $ echo 'client side corruption' > r7/.hg/largefiles/4cdac4d8b084d0b599525cf732437fb337d422a8
  $ rm "$USERCACHE/4cdac4d8b084d0b599525cf732437fb337d422a8"
  $ hg init empty
  $ hg serve -R empty -d -p $HGPORT1 --pid-file hg.pid \
  >   --config 'web.allow_push=*' --config web.push_ssl=False
  $ cat hg.pid >> $DAEMON_PIDS
  $ hg push -R r7 http://localhost:$HGPORT1
  pushing to http://localhost:$HGPORT1/
  searching for changes
  remote: largefiles: failed to put 4cdac4d8b084d0b599525cf732437fb337d422a8 into store: largefile contents do not match hash
  abort: remotestore: could not put $TESTTMP/r7/.hg/largefiles/4cdac4d8b084d0b599525cf732437fb337d422a8 to remote store http://localhost:$HGPORT1/ (glob)
  [255]
  $ mv 4cdac4d8b084d0b599525cf732437fb337d422a8 r7/.hg/largefiles/4cdac4d8b084d0b599525cf732437fb337d422a8
Push of file that exists on server but is corrupted - magic healing would be nice ... but too magic
  $ echo "server side corruption" > empty/.hg/largefiles/4cdac4d8b084d0b599525cf732437fb337d422a8
  $ hg push -R r7 http://localhost:$HGPORT1
  pushing to http://localhost:$HGPORT1/
  searching for changes
  remote: adding changesets
  remote: adding manifests
  remote: adding file changes
  remote: added 2 changesets with 2 changes to 2 files
  $ cat empty/.hg/largefiles/4cdac4d8b084d0b599525cf732437fb337d422a8
  server side corruption
  $ rm -rf empty

Push a largefiles repository to a served empty repository
  $ hg init r8
  $ echo c3 > r8/f1
  $ hg add --large r8/f1 -R r8
  $ hg commit -m "m1" -R r8
  Invoking status precommit hook
  A f1
  $ hg init empty
  $ hg serve -R empty -d -p $HGPORT2 --pid-file hg.pid \
  >   --config 'web.allow_push=*' --config web.push_ssl=False
  $ cat hg.pid >> $DAEMON_PIDS
  $ rm "${USERCACHE}"/*
  $ hg push -R r8 http://localhost:$HGPORT2/#default
  pushing to http://localhost:$HGPORT2/
  searching for changes
  remote: adding changesets
  remote: adding manifests
  remote: adding file changes
  remote: added 1 changesets with 1 changes to 1 files
  $ [ -f "${USERCACHE}"/02a439e5c31c526465ab1a0ca1f431f76b827b90 ]
  $ [ -f empty/.hg/largefiles/02a439e5c31c526465ab1a0ca1f431f76b827b90 ]

Clone over http, no largefiles pulled on clone.

  $ hg clone http://localhost:$HGPORT2/#default http-clone -U
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files

Archive contains largefiles
  >>> import urllib2, os
  >>> u = 'http://localhost:%s/archive/default.zip' % os.environ['HGPORT2']
  >>> with open('archive.zip', 'w') as f:
  ...     f.write(urllib2.urlopen(u).read())
  $ unzip -t archive.zip
  Archive:  archive.zip
      testing: empty-default/.hg_archival.txt*OK (glob)
      testing: empty-default/f1*OK (glob)
  No errors detected in compressed data of archive.zip.

test 'verify' with remotestore:

  $ rm "${USERCACHE}"/02a439e5c31c526465ab1a0ca1f431f76b827b90
  $ mv empty/.hg/largefiles/02a439e5c31c526465ab1a0ca1f431f76b827b90 .
  $ hg -R http-clone verify --large --lfa
  checking changesets
  checking manifests
  crosschecking files in changesets and manifests
  checking files
  1 files, 1 changesets, 1 total revisions
  searching 1 changesets for largefiles
  changeset 0:cf03e5bb9936: f1 missing
  verified existence of 1 revisions of 1 largefiles
  [1]
  $ mv 02a439e5c31c526465ab1a0ca1f431f76b827b90 empty/.hg/largefiles/
  $ hg -R http-clone -q verify --large --lfa

largefiles pulled on update - a largefile missing on the server:
  $ mv empty/.hg/largefiles/02a439e5c31c526465ab1a0ca1f431f76b827b90 .
  $ hg -R http-clone up --config largefiles.usercache=http-clone-usercache
  getting changed largefiles
  f1: largefile 02a439e5c31c526465ab1a0ca1f431f76b827b90 not available from http://localhost:$HGPORT2/
  0 largefiles updated, 0 removed
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ hg -R http-clone st
  ! f1
  $ hg -R http-clone up -Cqr null

largefiles pulled on update - a largefile corrupted on the server:
  $ echo corruption > empty/.hg/largefiles/02a439e5c31c526465ab1a0ca1f431f76b827b90
  $ hg -R http-clone up --config largefiles.usercache=http-clone-usercache
  getting changed largefiles
  f1: data corruption (expected 02a439e5c31c526465ab1a0ca1f431f76b827b90, got 6a7bb2556144babe3899b25e5428123735bb1e27)
  0 largefiles updated, 0 removed
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ hg -R http-clone st
  ! f1
  $ [ ! -f http-clone/.hg/largefiles/02a439e5c31c526465ab1a0ca1f431f76b827b90 ]
  $ [ ! -f http-clone/f1 ]
  $ [ ! -f http-clone-usercache ]
  $ hg -R http-clone verify --large --lfc
  checking changesets
  checking manifests
  crosschecking files in changesets and manifests
  checking files
  1 files, 1 changesets, 1 total revisions
  searching 1 changesets for largefiles
  verified contents of 1 revisions of 1 largefiles
  $ hg -R http-clone up -Cqr null

largefiles pulled on update - no server side problems:
  $ mv 02a439e5c31c526465ab1a0ca1f431f76b827b90 empty/.hg/largefiles/
  $ hg -R http-clone --debug up --config largefiles.usercache=http-clone-usercache --config progress.debug=true
  resolving manifests
   branchmerge: False, force: False, partial: False
   ancestor: 000000000000, local: 000000000000+, remote: cf03e5bb9936
   .hglf/f1: remote created -> g
  getting .hglf/f1
  updating: .hglf/f1 1/1 files (100.00%)
  getting changed largefiles
  using http://localhost:$HGPORT2/
  sending capabilities command
  sending batch command
  getting largefiles: 0/1 files (0.00%)
  getting f1:02a439e5c31c526465ab1a0ca1f431f76b827b90
  sending getlfile command
  found 02a439e5c31c526465ab1a0ca1f431f76b827b90 in store
  1 largefiles updated, 0 removed
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved

  $ ls http-clone-usercache/*
  http-clone-usercache/02a439e5c31c526465ab1a0ca1f431f76b827b90

  $ rm -rf empty http-clone*

used all HGPORTs, kill all daemons
  $ killdaemons.py

largefiles should batch verify remote calls

  $ hg init batchverifymain
  $ cd batchverifymain
  $ echo "aaa" >> a
  $ hg add --large a
  $ hg commit -m "a"
  Invoking status precommit hook
  A a
  $ echo "bbb" >> b
  $ hg add --large b
  $ hg commit -m "b"
  Invoking status precommit hook
  A b
  $ cd ..
  $ hg serve -R batchverifymain -d -p $HGPORT --pid-file hg.pid \
  > -A access.log
  $ cat hg.pid >> $DAEMON_PIDS
  $ hg clone --noupdate http://localhost:$HGPORT batchverifyclone
  requesting all changes
  adding changesets
  adding manifests
  adding file changes
  added 2 changesets with 2 changes to 2 files
  $ hg -R batchverifyclone verify --large --lfa
  checking changesets
  checking manifests
  crosschecking files in changesets and manifests
  checking files
  2 files, 2 changesets, 2 total revisions
  searching 2 changesets for largefiles
  verified existence of 2 revisions of 2 largefiles
  $ tail -1 access.log
  127.0.0.1 - - [*] "GET /?cmd=batch HTTP/1.1" 200 - x-hgarg-1:cmds=statlfile+sha%3D972a1a11f19934401291cc99117ec614933374ce%3Bstatlfile+sha%3Dc801c9cfe94400963fcb683246217d5db77f9a9a x-hgproto-1:0.1 0.2 comp=*zlib,none,bzip2 (glob)
  $ hg -R batchverifyclone update
  getting changed largefiles
  2 largefiles updated, 0 removed
  2 files updated, 0 files merged, 0 files removed, 0 files unresolved

Clear log file before next test

  $ printf "" > access.log

Verify should check file on remote server only when file is not
available locally.

  $ echo "ccc" >> batchverifymain/c
  $ hg -R batchverifymain status
  ? c
  $ hg -R batchverifymain add --large batchverifymain/c
  $ hg -R batchverifymain commit -m "c"
  Invoking status precommit hook
  A c
  $ hg -R batchverifyclone pull
  pulling from http://localhost:$HGPORT/
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files
  (run 'hg update' to get a working copy)
  $ hg -R batchverifyclone verify --lfa
  checking changesets
  checking manifests
  crosschecking files in changesets and manifests
  checking files
  3 files, 3 changesets, 3 total revisions
  searching 3 changesets for largefiles
  verified existence of 3 revisions of 3 largefiles
  $ tail -1 access.log
  127.0.0.1 - - [*] "GET /?cmd=batch HTTP/1.1" 200 - x-hgarg-1:cmds=statlfile+sha%3Dc8559c3c9cfb42131794b7d8009230403b9b454c x-hgproto-1:0.1 0.2 comp=*zlib,none,bzip2 (glob)

  $ killdaemons.py

largefiles should not ask for password again after successful authorization

  $ hg init credentialmain
  $ cd credentialmain
  $ echo "aaa" >> a
  $ hg add --large a
  $ hg commit -m "a"
  Invoking status precommit hook
  A a

Before running server clear the user cache to force clone to download
a large file from the server rather than to get it from the cache

  $ rm "${USERCACHE}"/*

  $ cd ..
  $ cat << EOT > userpass.py
  > import base64
  > from mercurial.hgweb import common
  > def perform_authentication(hgweb, req, op):
  >     auth = req.env.get('HTTP_AUTHORIZATION')
  >     if not auth:
  >         raise common.ErrorResponse(common.HTTP_UNAUTHORIZED, 'who',
  >                 [('WWW-Authenticate', 'Basic Realm="mercurial"')])
  >     if base64.b64decode(auth.split()[1]).split(':', 1) != ['user', 'pass']:
  >         raise common.ErrorResponse(common.HTTP_FORBIDDEN, 'no')
  > def extsetup():
  >     common.permhooks.insert(0, perform_authentication)
  > EOT
  $ hg serve --config extensions.x=userpass.py -R credentialmain \
  >          -d -p $HGPORT --pid-file hg.pid -A access.log
  $ cat hg.pid >> $DAEMON_PIDS
  $ cat << EOF > get_pass.py
  > import getpass
  > def newgetpass(arg):
  >   return "pass"
  > getpass.getpass = newgetpass
  > EOF
  $ hg clone --config ui.interactive=true --config extensions.getpass=get_pass.py \
  >          http://user@localhost:$HGPORT credentialclone
  requesting all changes
  http authorization required for http://localhost:$HGPORT/
  realm: mercurial
  user: user
  password: adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files
  updating to branch default
  getting changed largefiles
  1 largefiles updated, 0 removed
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved

  $ rm hg.pid access.log
  $ killdaemons.py

#endif