Mercurial > hg
view contrib/hgperf @ 30442:41a8106789ca
util: implement zstd compression engine
Now that zstd is vendored and being built (in some configurations), we
can implement a compression engine for zstd!
The zstd engine is a little different from existing engines. Because
it may not always be present, we have to defer load the module in case
importing it fails. We facilitate this via a cached property that holds
a reference to the module or None. The "available" method is
implemented to reflect reality.
The zstd engine declares its ability to handle bundles using the
"zstd" human name and the "ZS" internal name. The latter was chosen
because internal names are 2 characters (by only convention I think)
and "ZS" seems reasonable.
The engine, like others, supports specifying the compression level.
However, there are no consumers of this API that yet pass in that
argument. I have plans to change that, so stay tuned.
Since all we need to do to support bundle generation with a new
compression engine is implement and register the compression engine,
bundle generation with zstd "just works!" Tests demonstrating this
have been added.
How does performance of zstd for bundle generation compare? On the
mozilla-unified repo, `hg bundle --all -t <engine>-v2` yields the
following on my i7-6700K on Linux:
engine CPU time bundle size vs orig size throughput
none 97.0s 4,054,405,584 100.0% 41.8 MB/s
bzip2 (l=9) 393.6s 975,343,098 24.0% 10.3 MB/s
gzip (l=6) 184.0s 1,140,533,074 28.1% 22.0 MB/s
zstd (l=1) 108.2s 1,119,434,718 27.6% 37.5 MB/s
zstd (l=2) 111.3s 1,078,328,002 26.6% 36.4 MB/s
zstd (l=3) 113.7s 1,011,823,727 25.0% 35.7 MB/s
zstd (l=4) 116.0s 1,008,965,888 24.9% 35.0 MB/s
zstd (l=5) 121.0s 977,203,148 24.1% 33.5 MB/s
zstd (l=6) 131.7s 927,360,198 22.9% 30.8 MB/s
zstd (l=7) 139.0s 912,808,505 22.5% 29.2 MB/s
zstd (l=12) 198.1s 854,527,714 21.1% 20.5 MB/s
zstd (l=18) 681.6s 789,750,690 19.5% 5.9 MB/s
On compression, zstd for bundle generation delivers:
* better compression than gzip with significantly less CPU utilization
* better than bzip2 compression ratios while still being significantly
faster than gzip
* ability to aggressively tune compression level to achieve
significantly smaller bundles
That last point is important. With clone bundles, a server can
pre-generate a bundle file, upload it to a static file server, and
redirect clients to transparently download it during clone. The server
could choose to produce a zstd bundle with the highest compression
settings possible. This would take a very long time - a magnitude
longer than a typical zstd bundle generation - but the result would
be hundreds of megabytes smaller! For the clone volume we do at
Mozilla, this could translate to petabytes of bandwidth savings
per year and faster clones (due to smaller transfer size).
I don't have detailed numbers to report on decompression. However,
zstd decompression is fast: >1 GB/s output throughput on this machine,
even through the Python bindings. And it can do that regardless of the
compression level of the input. By the time you have enough data to
worry about overhead of decompression, you have plenty of other things
to worry about performance wise.
zstd is wins all around. I can't wait to implement support for it
on the wire protocol and in revlogs.
author | Gregory Szorc <gregory.szorc@gmail.com> |
---|---|
date | Fri, 11 Nov 2016 01:10:07 -0800 |
parents | 377a111d1cd2 |
children | 22fbca1d11ed |
line wrap: on
line source
#!/usr/bin/env python # # hgperf - measure performance of Mercurial commands # # Copyright 2014 Matt Mackall <mpm@selenic.com> # # This software may be used and distributed according to the terms of the # GNU General Public License version 2 or any later version. '''measure performance of Mercurial commands Using ``hgperf`` instead of ``hg`` measures performance of the target Mercurial command. For example, the execution below measures performance of :hg:`heads --topo`:: $ hgperf heads --topo All command output via ``ui`` is suppressed, and just measurement result is displayed: see also "perf" extension in "contrib". Costs of processing before dispatching to the command function like below are not measured:: - parsing command line (e.g. option validity check) - reading configuration files in But ``pre-`` and ``post-`` hook invocation for the target command is measured, even though these are invoked before or after dispatching to the command function, because these may be required to repeat execution of the target command correctly. ''' import os import sys libdir = '@LIBDIR@' if libdir != '@' 'LIBDIR' '@': if not os.path.isabs(libdir): libdir = os.path.join(os.path.dirname(os.path.realpath(__file__)), libdir) libdir = os.path.abspath(libdir) sys.path.insert(0, libdir) # enable importing on demand to reduce startup time try: from mercurial import demandimport; demandimport.enable() except ImportError: import sys sys.stderr.write("abort: couldn't find mercurial libraries in [%s]\n" % ' '.join(sys.path)) sys.stderr.write("(check your install and PYTHONPATH)\n") sys.exit(-1) import mercurial.util import mercurial.dispatch import time def timer(func, title=None): results = [] begin = time.time() count = 0 while True: ostart = os.times() cstart = time.time() r = func() cstop = time.time() ostop = os.times() count += 1 a, b = ostart, ostop results.append((cstop - cstart, b[0] - a[0], b[1]-a[1])) if cstop - begin > 3 and count >= 100: break if cstop - begin > 10 and count >= 3: break if title: sys.stderr.write("! %s\n" % title) if r: sys.stderr.write("! result: %s\n" % r) m = min(results) sys.stderr.write("! wall %f comb %f user %f sys %f (best of %d)\n" % (m[0], m[1] + m[2], m[1], m[2], count)) orgruncommand = mercurial.dispatch.runcommand def runcommand(lui, repo, cmd, fullargs, ui, options, d, cmdpats, cmdoptions): ui.pushbuffer() lui.pushbuffer() timer(lambda : orgruncommand(lui, repo, cmd, fullargs, ui, options, d, cmdpats, cmdoptions)) ui.popbuffer() lui.popbuffer() mercurial.dispatch.runcommand = runcommand for fp in (sys.stdin, sys.stdout, sys.stderr): mercurial.util.setbinary(fp) mercurial.dispatch.run()