tests/test-contrib-perf.t
author Gregory Szorc <gregory.szorc@gmail.com>
Mon, 26 Mar 2018 11:00:16 -0700
changeset 37288 9bfcbe4f4745
parent 35901 ed939545edd0
child 37338 cbc4425e81b5
permissions -rw-r--r--
wireproto: add streams to frame-based protocol Previously, the frame-based protocol was just a series of frames, with each frame associated with a request ID. In order to scale the protocol, we'll want to enable the use of compression. While it is possible to enable compression at the socket/pipe level, this has its disadvantages. The big one is it undermines the point of frames being standalone, atomic units that can be read and written: if you add compression above the framing protocol, you are back to having a stream-based protocol as opposed to something frame-based. So in order to preserve frames, compression needs to occur at the frame payload level. Compressing each frame's payload individually will limit compression ratios because the window size of the compressor will be limited by the max frame size, which is 32-64kb as currently defined. It will also add CPU overhead, as it is more efficient for compressors to operate on fewer, larger blocks of data than more, smaller blocks. So compressing each frame independently is out. This means we need to compress each frame's payload as if it is part of a larger stream. The simplest approach is to have 1 stream per connection. This could certainly work. However, it has disadvantages (documented below). We could also have 1 stream per RPC/command invocation. (This is the model HTTP/2 goes with.) This also has disadvantages. The main disadvantage to one global stream is that it has the very real potential to create CPU bottlenecks doing compression. Networks are only getting faster and the performance of single CPU cores has been relatively flat. Newer compression formats like zstandard offer better CPU cycle efficiency than predecessors like zlib. But it still all too common to saturate your CPU with compression overhead long before you saturate the network pipe. The main disadvantage with streams per request is that you can't reap the benefits of the compression context for multiple requests. For example, if you send 1000 RPC requests (or HTTP/2 requests for that matter), the response to each would have its own compression context. The overall size of the raw responses would be larger because compression contexts wouldn't be able to reference data from another request or response. The approach for streams as implemented in this commit is to support N streams per connection and for streams to potentially span requests and responses. As explained by the added internals docs, this facilitates servers and clients delegating independent streams and compression to independent threads / CPU cores. This helps alleviate the CPU bottleneck of compression. This design also allows compression contexts to be reused across requests/responses. This can result in improved compression ratios and less overhead for compressors and decompressors having to build new contexts. Another feature that was defined was the ability for individual frames within a stream to declare whether that individual frame's payload uses the content encoding (read: compression) defined by the stream. The idea here is that some servers may serve data from a combination of caches and dynamic resolution. Data coming from caches may be pre-compressed. We want to facilitate servers being able to essentially stream bytes from caches to the wire with minimal overhead. Being able to mix and match with frames are compressed within a stream enables these types of advanced server functionality. This commit defines the new streams mechanism. Basic code for supporting streams in frames has been added. But that code is seriously lacking and doesn't fully conform to the defined protocol. For example, we don't close any streams. And support for content encoding within streams is not yet implemented. The change was rather invasive and I didn't think it would be reasonable to implement the entire feature in a single commit. For the record, I would have loved to reuse an existing multiplexing protocol to build the new wire protocol on top of. However, I couldn't find a protocol that offers the performance and scaling characteristics that I desired. Namely, it should support multiple compression contexts to facilitate scaling out to multiple CPU cores and compression contexts should be able to live longer than single RPC requests. HTTP/2 *almost* fits the bill. But the semantics of HTTP message exchange state that streams can only live for a single request-response. We /could/ tunnel on top of HTTP/2 streams and frames with HEADER and DATA frames. But there's no guarantee that HTTP/2 libraries and proxies would allow us to use HTTP/2 streams and frames without the HTTP message exchange semantics defined in RFC 7540 Section 8. Other RPC protocols like gRPC tunnel are built on top of HTTP/2 and thus preserve its semantics of stream per RPC invocation. Even QUIC does this. We could attempt to invent a higher-level stream that spans HTTP/2 streams. But this would be violating HTTP/2 because there is no guarantee that HTTP/2 streams are routed to the same server. The best we can do - which is what this protocol does - is shoehorn all request and response data into a single HTTP message and create streams within. At that point, we've defined a Content-Type in HTTP parlance. It just so happens our media type can also work as a standalone, stream-based protocol, without leaning on HTTP or similar protocol. Differential Revision: https://phab.mercurial-scm.org/D2907

#require test-repo

Set vars:

  $ . "$TESTDIR/helpers-testrepo.sh"
  $ CONTRIBDIR="$TESTDIR/../contrib"

Prepare repo:

  $ hg init

  $ echo this is file a > a
  $ hg add a
  $ hg commit -m first

  $ echo adding to file a >> a
  $ hg commit -m second

  $ echo adding more to file a >> a
  $ hg commit -m third

  $ hg up -r 0
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ echo merge-this >> a
  $ hg commit -m merge-able
  created new head

  $ hg up -r 2
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved

perfstatus

  $ cat >> $HGRCPATH << EOF
  > [extensions]
  > perfstatusext=$CONTRIBDIR/perf.py
  > [perf]
  > presleep=0
  > stub=on
  > parentscount=1
  > EOF
  $ hg help perfstatusext
  perfstatusext extension - helper extension to measure performance
  
  list of commands:
  
   perfaddremove
                 (no help text available)
   perfancestors
                 (no help text available)
   perfancestorset
                 (no help text available)
   perfannotate  (no help text available)
   perfbdiff     benchmark a bdiff between revisions
   perfbookmarks
                 benchmark parsing bookmarks from disk to memory
   perfbranchmap
                 benchmark the update of a branchmap
   perfbundleread
                 Benchmark reading of bundle files.
   perfcca       (no help text available)
   perfchangegroupchangelog
                 Benchmark producing a changelog group for a changegroup.
   perfchangeset
                 (no help text available)
   perfctxfiles  (no help text available)
   perfdiffwd    Profile diff of working directory changes
   perfdirfoldmap
                 (no help text available)
   perfdirs      (no help text available)
   perfdirstate  (no help text available)
   perfdirstatedirs
                 (no help text available)
   perfdirstatefoldmap
                 (no help text available)
   perfdirstatewrite
                 (no help text available)
   perffncacheencode
                 (no help text available)
   perffncacheload
                 (no help text available)
   perffncachewrite
                 (no help text available)
   perfheads     (no help text available)
   perfindex     (no help text available)
   perfloadmarkers
                 benchmark the time to parse the on-disk markers for a repo
   perflog       (no help text available)
   perflookup    (no help text available)
   perflrucachedict
                 (no help text available)
   perfmanifest  (no help text available)
   perfmergecalculate
                 (no help text available)
   perfmoonwalk  benchmark walking the changelog backwards
   perfnodelookup
                 (no help text available)
   perfparents   (no help text available)
   perfpathcopies
                 (no help text available)
   perfphases    benchmark phasesets computation
   perfrawfiles  (no help text available)
   perfrevlogchunks
                 Benchmark operations on revlog chunks.
   perfrevlogindex
                 Benchmark operations against a revlog index.
   perfrevlogrevision
                 Benchmark obtaining a revlog revision.
   perfrevlogrevisions
                 Benchmark reading a series of revisions from a revlog.
   perfrevrange  (no help text available)
   perfrevset    benchmark the execution time of a revset
   perfstartup   (no help text available)
   perfstatus    (no help text available)
   perftags      (no help text available)
   perftemplating
                 (no help text available)
   perfunidiff   benchmark a unified diff between revisions
   perfvolatilesets
                 benchmark the computation of various volatile set
   perfwalk      (no help text available)
   perfwrite     microbenchmark ui.write
  
  (use 'hg help -v perfstatusext' to show built-in aliases and global options)
  $ hg perfaddremove
  $ hg perfancestors
  $ hg perfancestorset 2
  $ hg perfannotate a
  $ hg perfbdiff -c 1
  $ hg perfbdiff --alldata 1
  $ hg perfunidiff -c 1
  $ hg perfunidiff --alldata 1
  $ hg perfbookmarks
  $ hg perfbranchmap
  $ hg perfcca
  $ hg perfchangegroupchangelog
  $ hg perfchangeset 2
  $ hg perfctxfiles 2
  $ hg perfdiffwd
  $ hg perfdirfoldmap
  $ hg perfdirs
  $ hg perfdirstate
  $ hg perfdirstatedirs
  $ hg perfdirstatefoldmap
  $ hg perfdirstatewrite
  $ hg perffncacheencode
  $ hg perffncacheload
  $ hg perffncachewrite
  $ hg perfheads
  $ hg perfindex
  $ hg perfloadmarkers
  $ hg perflog
  $ hg perflookup 2
  $ hg perflrucache
  $ hg perfmanifest 2
  $ hg perfmergecalculate -r 3
  $ hg perfmoonwalk
  $ hg perfnodelookup 2
  $ hg perfpathcopies 1 2
  $ hg perfrawfiles 2
  $ hg perfrevlogindex -c
  $ hg perfrevlogrevisions .hg/store/data/a.i
  $ hg perfrevlogrevision -m 0
  $ hg perfrevlogchunks -c
  $ hg perfrevrange
  $ hg perfrevset 'all()'
  $ hg perfstartup
  $ hg perfstatus
  $ hg perftags
  $ hg perftemplating
  $ hg perfvolatilesets
  $ hg perfwalk
  $ hg perfparents

Check perf.py for historical portability

  $ cd "$TESTDIR/.."

  $ (testrepohg files -r 1.2 glob:mercurial/*.c glob:mercurial/*.py;
  >  testrepohg files -r tip glob:mercurial/*.c glob:mercurial/*.py) |
  > "$TESTDIR"/check-perf-code.py contrib/perf.py
  contrib/perf.py:\d+: (re)
   >     from mercurial import (
   import newer module separately in try clause for early Mercurial
  [1]