##// END OF EJS Templates
compression: introduce a `storage.revlog.zlib.level` configuration...
compression: introduce a `storage.revlog.zlib.level` configuration This option control the zlib compression level used when compression revlog chunk. This is also a good excuse to pave the way for a similar configuration option for the zstd compression engine. Having a dedicated option for each compression algorithm is useful because they don't support the same range of values. Using a higher zlib compression impact CPU consumption at compression time, but does not directly affected decompression time. However dealing with small compressed chunk can directly help decompression and indirectly help other revlog logic. I ran some basic test on repositories using different level. I am using the mercurial, pypy, netbeans and mozilla-central clone from our benchmark suite. All tested repository use sparse-revlog and got all their delta recomputed. The different compression level has a small effect on the repository size (about 10% variation in the total range). My quick analysis is that revlog mostly store small delta, that are not affected by the compression level much. So the variation probably mostly comes from better compression of the snapshots revisions, and snapshot revision only represent a small portion of the repository content. I also made some basic timings measurements. The "read" timings are gathered using simple run of `hg perfrevlogrevisions`, the "write" timings using `hg perfrevlogwrite` (restricted to the last 5000 revisions for netbeans and mozilla central). The timings are gathered on a generic machine, (not one of our performance locked machine), so small variation might not be meaningful. However large trend remains relevant. Keep in mind that these numbers are not pure compression/decompression time. They also involve the full revlog logic. In particular the difference in chunk size has an impact on the delta chain structure, affecting performance when writing or reading them. On read/write performance, the compression level has a bigger impact. Counter-intuitively, the higher compression levels improve "write" performance for the large repositories in our tested setting. Maybe because the last 5000 delta chain end up having a very different shape in this specific spot? Or maybe because of a more general trend of better delta chains thanks to the smaller chunk and snapshot. This series does not intend to change the default compression level. However, these result call for a deeper analysis of this performance difference in the future. Full data ========= repo level .hg/store size 00manifest.d read write ---------------------------------------------------------------- mercurial 1 49,402,813 5,963,475 0.170159 53.250304 mercurial 6 47,197,397 5,875,730 0.182820 56.264320 mercurial 9 47,121,596 5,849,781 0.189219 56.293612 pypy 1 370,830,572 28,462,425 2.679217 460.721984 pypy 6 340,112,317 27,648,747 2.768691 467.537158 pypy 9 338,360,736 27,639,003 2.763495 476.589918 netbeans 1 1,281,847,810 165,495,457 122.477027 520.560316 netbeans 6 1,205,284,353 159,161,207 139.876147 715.930400 netbeans 9 1,197,135,671 155,034,586 141.620281 678.297064 mozilla 1 2,775,497,186 298,527,987 147.867662 751.263721 mozilla 6 2,596,856,420 286,597,671 170.572118 987.056093 mozilla 9 2,587,542,494 287,018,264 163.622338 739.803002

File last commit:

r41401:876494fd default
r42210:1fac9b93 default
Show More
protocol.py
227 lines | 7.6 KiB | text/x-python | PythonLexer
# Copyright 2016-present Facebook. All Rights Reserved.
#
# protocol: logic for a server providing fastannotate support
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
from __future__ import absolute_import
import contextlib
import os
from mercurial.i18n import _
from mercurial import (
error,
extensions,
hg,
util,
wireprotov1peer,
wireprotov1server,
)
from . import context
# common
def _getmaster(ui):
"""get the mainbranch, and enforce it is set"""
master = ui.config('fastannotate', 'mainbranch')
if not master:
raise error.Abort(_('fastannotate.mainbranch is required '
'for both the client and the server'))
return master
# server-side
def _capabilities(orig, repo, proto):
result = orig(repo, proto)
result.append('getannotate')
return result
def _getannotate(repo, proto, path, lastnode):
# output:
# FILE := vfspath + '\0' + str(size) + '\0' + content
# OUTPUT := '' | FILE + OUTPUT
result = ''
buildondemand = repo.ui.configbool('fastannotate', 'serverbuildondemand',
True)
with context.annotatecontext(repo, path) as actx:
if buildondemand:
# update before responding to the client
master = _getmaster(repo.ui)
try:
if not actx.isuptodate(master):
actx.annotate(master, master)
except Exception:
# non-fast-forward move or corrupted. rebuild automically.
actx.rebuild()
try:
actx.annotate(master, master)
except Exception:
actx.rebuild() # delete files
finally:
# although the "with" context will also do a close/flush, we
# need to do it early so we can send the correct respond to
# client.
actx.close()
# send back the full content of revmap and linelog, in the future we
# may want to do some rsync-like fancy updating.
# the lastnode check is not necessary if the client and the server
# agree where the main branch is.
if actx.lastnode != lastnode:
for p in [actx.revmappath, actx.linelogpath]:
if not os.path.exists(p):
continue
with open(p, 'rb') as f:
content = f.read()
vfsbaselen = len(repo.vfs.base + '/')
relpath = p[vfsbaselen:]
result += '%s\0%d\0%s' % (relpath, len(content), content)
return result
def _registerwireprotocommand():
if 'getannotate' in wireprotov1server.commands:
return
wireprotov1server.wireprotocommand(
'getannotate', 'path lastnode')(_getannotate)
def serveruisetup(ui):
_registerwireprotocommand()
extensions.wrapfunction(wireprotov1server, '_capabilities', _capabilities)
# client-side
def _parseresponse(payload):
result = {}
i = 0
l = len(payload) - 1
state = 0 # 0: vfspath, 1: size
vfspath = size = ''
while i < l:
ch = payload[i:i + 1]
if ch == '\0':
if state == 1:
result[vfspath] = payload[i + 1:i + 1 + int(size)]
i += int(size)
state = 0
vfspath = size = ''
elif state == 0:
state = 1
else:
if state == 1:
size += ch
elif state == 0:
vfspath += ch
i += 1
return result
def peersetup(ui, peer):
class fastannotatepeer(peer.__class__):
@wireprotov1peer.batchable
def getannotate(self, path, lastnode=None):
if not self.capable('getannotate'):
ui.warn(_('remote peer cannot provide annotate cache\n'))
yield None, None
else:
args = {'path': path, 'lastnode': lastnode or ''}
f = wireprotov1peer.future()
yield args, f
yield _parseresponse(f.value)
peer.__class__ = fastannotatepeer
@contextlib.contextmanager
def annotatepeer(repo):
ui = repo.ui
remotepath = ui.expandpath(
ui.config('fastannotate', 'remotepath', 'default'))
peer = hg.peer(ui, {}, remotepath)
try:
yield peer
finally:
peer.close()
def clientfetch(repo, paths, lastnodemap=None, peer=None):
"""download annotate cache from the server for paths"""
if not paths:
return
if peer is None:
with annotatepeer(repo) as peer:
return clientfetch(repo, paths, lastnodemap, peer)
if lastnodemap is None:
lastnodemap = {}
ui = repo.ui
results = []
with peer.commandexecutor() as batcher:
ui.debug('fastannotate: requesting %d files\n' % len(paths))
for p in paths:
results.append(batcher.callcommand(
'getannotate',
{'path': p, 'lastnode':lastnodemap.get(p)}))
for result in results:
r = result.result()
# TODO: pconvert these paths on the server?
r = {util.pconvert(p): v for p, v in r.iteritems()}
for path in sorted(r):
# ignore malicious paths
if (not path.startswith('fastannotate/')
or '/../' in (path + '/')):
ui.debug('fastannotate: ignored malicious path %s\n' % path)
continue
content = r[path]
if ui.debugflag:
ui.debug('fastannotate: writing %d bytes to %s\n'
% (len(content), path))
repo.vfs.makedirs(os.path.dirname(path))
with repo.vfs(path, 'wb') as f:
f.write(content)
def _filterfetchpaths(repo, paths):
"""return a subset of paths whose history is long and need to fetch linelog
from the server. works with remotefilelog and non-remotefilelog repos.
"""
threshold = repo.ui.configint('fastannotate', 'clientfetchthreshold', 10)
if threshold <= 0:
return paths
result = []
for path in paths:
try:
if len(repo.file(path)) >= threshold:
result.append(path)
except Exception: # file not found etc.
result.append(path)
return result
def localreposetup(ui, repo):
class fastannotaterepo(repo.__class__):
def prefetchfastannotate(self, paths, peer=None):
master = _getmaster(self.ui)
needupdatepaths = []
lastnodemap = {}
try:
for path in _filterfetchpaths(self, paths):
with context.annotatecontext(self, path) as actx:
if not actx.isuptodate(master, strict=False):
needupdatepaths.append(path)
lastnodemap[path] = actx.lastnode
if needupdatepaths:
clientfetch(self, needupdatepaths, lastnodemap, peer)
except Exception as ex:
# could be directory not writable or so, not fatal
self.ui.debug('fastannotate: prefetch failed: %r\n' % ex)
repo.__class__ = fastannotaterepo
def clientreposetup(ui, repo):
_registerwireprotocommand()
if repo.local():
localreposetup(ui, repo)
# TODO: this mutates global state, but only if at least one repo
# has the extension enabled. This is probably bad for hgweb.
if peersetup not in hg.wirepeersetupfuncs:
hg.wirepeersetupfuncs.append(peersetup)