view tests/test-mq-git.t @ 40021:c537144fdbef

wireprotov2: support response caching

One of the things I've learned from managing VCS servers over the
years is that they are hard to scale. It is well known that some
companies have very beefy (read: very expensive) servers to power
their VCS needs. It is also known that specialized servers for
various VCS exist in order to facilitate scaling servers. (Mercurial
is in this boat.)

One of the aspects that make a VCS server hard to scale is the
high CPU load incurred by constant client clone/pull operations.

To alleviate the scaling pain associated with data retrieval
operations, I want to integrate caching into the Mercurial wire
protocol server as robustly as possible such that servers can
aggressively cache responses and defer as much server load as
possible.

This commit represents the initial implementation of a general
caching layer in wire protocol version 2.

We define a new interface and behavior for a wire protocol cacher
in repository.py. (This is probably where a reviewer should look
first to understand what is going on.)

The bulk of the added code is in wireprotov2server.py, where we
define how a command can opt in to being cached and integrate
caching into command dispatching.

From a very high-level:

* A command can declare itself as cacheable by providing a callable
  that can be used to derive a cache key.
* At dispatch time, if a command is cacheable, we attempt to
  construct a cacher and use it for serving the request and/or
  caching the request.
* The dispatch layer handles the bulk of the business logic for
  caching, making cachers mostly "dumb content stores."
* The mechanism for invalidating cached entries (one of the harder
  parts about caching in general) is by varying the cache key when
  state changes. As such, cachers don't need to be concerned with
  cache invalidation.

Initially, we've hooked up support for caching "manifestdata" and
"filedata" commands. These are the simplest to cache, as they should
be immutable over time. Caching of commands related to changeset
data is a bit harder (because cache validation is impacted by
changes to bookmarks, phases, etc). This will be implemented later.
(Strictly speaking, censoring a file should invalidate caches. I've
added an inline TODO to track this edge case.)

To prove it works, this commit implements a test-only extension
providing in-memory caching backed by an lrucachedict. A new test
showing this extension behaving properly is added. FWIW, the
cacher is ~50 lines of code, demonstrating the relative ease with
which a cache can be added to a server.

While the test cacher is not suitable for production workloads, just
for kicks I performed a clone of just the changeset and manifest data
for the mozilla-unified repository. With a fully warmed cache (of just
the manifest data since changeset data is not cached), server-side
CPU usage dropped from ~73s to ~28s. That's pretty significant and
demonstrates the potential that response caching has on server
scalability!

Differential Revision: https://phab.mercurial-scm.org/D4773

# Test the plumbing of mq.git option
# Automatic upgrade itself is tested elsewhere.

  $ cat <<EOF >> $HGRCPATH
  > [extensions]
  > mq =
  > [diff]
  > nodates = 1
  > EOF

  $ hg init repo-auto
  $ cd repo-auto

git=auto: regular patch creation:

  $ echo a > a
  $ hg add a
  $ hg qnew -d '0 0' -f adda

  $ cat .hg/patches/adda
  # HG changeset patch
  # Date 0 0
  # Parent  0000000000000000000000000000000000000000
  
  diff -r 000000000000 -r ef8dafc9fa4c a
  --- /dev/null
  +++ b/a
  @@ -0,0 +1,1 @@
  +a

git=auto: git patch creation with copy:

  $ hg cp a b
  $ hg qnew -d '0 0' -f copy

  $ cat .hg/patches/copy
  # HG changeset patch
  # Date 0 0
  # Parent  ef8dafc9fa4caff80f6e243eb0171bcd60c455b4
  
  diff --git a/a b/b
  copy from a
  copy to b

git=auto: git patch when using --git:

  $ echo regular > regular
  $ hg add regular
  $ hg qnew -d '0 0' --git -f git

  $ cat .hg/patches/git
  # HG changeset patch
  # Date 0 0
  # Parent  99586d5f048c399e20f81cee41fbb3809c0e735d
  
  diff --git a/regular b/regular
  new file mode 100644
  --- /dev/null
  +++ b/regular
  @@ -0,0 +1,1 @@
  +regular

git=auto: regular patch after qrefresh without --git:

  $ hg qrefresh -d '0 0'

  $ cat .hg/patches/git
  # HG changeset patch
  # Date 0 0
  # Parent  99586d5f048c399e20f81cee41fbb3809c0e735d
  
  diff -r 99586d5f048c regular
  --- /dev/null
  +++ b/regular
  @@ -0,0 +1,1 @@
  +regular

  $ cd ..

  $ hg init repo-keep
  $ cd repo-keep
  $ echo '[mq]' > .hg/hgrc
  $ echo 'git = KEEP' >> .hg/hgrc

git=keep: git patch with --git:

  $ echo a > a
  $ hg add a
  $ hg qnew -d '0 0' -f --git git

  $ cat .hg/patches/git
  # HG changeset patch
  # Date 0 0
  # Parent  0000000000000000000000000000000000000000
  
  diff --git a/a b/a
  new file mode 100644
  --- /dev/null
  +++ b/a
  @@ -0,0 +1,1 @@
  +a

git=keep: git patch after qrefresh without --git:

  $ echo a >> a
  $ hg qrefresh -d '0 0'

  $ cat .hg/patches/git
  # HG changeset patch
  # Date 0 0
  # Parent  0000000000000000000000000000000000000000
  
  diff --git a/a b/a
  new file mode 100644
  --- /dev/null
  +++ b/a
  @@ -0,0 +1,2 @@
  +a
  +a
  $ cd ..

  $ hg init repo-yes
  $ cd repo-yes
  $ echo '[mq]' > .hg/hgrc
  $ echo 'git = yes' >> .hg/hgrc

git=yes: git patch:

  $ echo a > a
  $ hg add a
  $ hg qnew -d '0 0' -f git

  $ cat .hg/patches/git
  # HG changeset patch
  # Date 0 0
  # Parent  0000000000000000000000000000000000000000
  
  diff --git a/a b/a
  new file mode 100644
  --- /dev/null
  +++ b/a
  @@ -0,0 +1,1 @@
  +a

git=yes: git patch after qrefresh:

  $ echo a >> a
  $ hg qrefresh -d '0 0'

  $ cat .hg/patches/git
  # HG changeset patch
  # Date 0 0
  # Parent  0000000000000000000000000000000000000000
  
  diff --git a/a b/a
  new file mode 100644
  --- /dev/null
  +++ b/a
  @@ -0,0 +1,2 @@
  +a
  +a
  $ cd ..

  $ hg init repo-no
  $ cd repo-no
  $ echo '[diff]' > .hg/hgrc
  $ echo 'git = True' >> .hg/hgrc
  $ echo '[mq]' > .hg/hgrc
  $ echo 'git = False' >> .hg/hgrc

git=no: regular patch with copy:

  $ echo a > a
  $ hg add a
  $ hg qnew -d '0 0' -f adda
  $ hg cp a b
  $ hg qnew -d '0 0' -f regular

  $ cat .hg/patches/regular
  # HG changeset patch
  # Date 0 0
  # Parent  ef8dafc9fa4caff80f6e243eb0171bcd60c455b4
  
  diff -r ef8dafc9fa4c -r a70404f79ba3 b
  --- /dev/null
  +++ b/b
  @@ -0,0 +1,1 @@
  +a

git=no: regular patch after qrefresh with copy:

  $ hg cp a c
  $ hg qrefresh -d '0 0'

  $ cat .hg/patches/regular
  # HG changeset patch
  # Date 0 0
  # Parent  ef8dafc9fa4caff80f6e243eb0171bcd60c455b4
  
  diff -r ef8dafc9fa4c b
  --- /dev/null
  +++ b/b
  @@ -0,0 +1,1 @@
  +a
  diff -r ef8dafc9fa4c c
  --- /dev/null
  +++ b/c
  @@ -0,0 +1,1 @@
  +a

Test how [diff] configuration influence and cause invalid or lossy patches:

  $ cat <<EOF >> .hg/hgrc
  > [mq]
  > git = AUTO
  > [diff]
  > nobinary = True
  > noprefix = True
  > showfunc = True
  > ignorews = True
  > ignorewsamount = True
  > ignoreblanklines = True
  > unified = 1
  > EOF

  $ echo ' a' > a
  $ hg qnew prepare -d '0 0'
  $ echo '  a' > a
  $ printf '\0' > b
  $ echo >> c
  $ hg qnew diff -d '0 0'

  $ cat .hg/patches/prepare
  # HG changeset patch
  # Date 0 0
  # Parent  cf0bfe72686a47d8d7d7b4529a3adb8b0b449a9f
  
  diff -r cf0bfe72686a -r fb9c4422b0f3 a
  --- a/a
  +++ b/a
  @@ -1,1 +1,1 @@
  -a
  + a
  $ cat .hg/patches/diff
  # HG changeset patch
  # Date 0 0
  # Parent  fb9c4422b0f37dd576522dd9a3f99b825c177efe
  
  diff --git a/a b/a
  --- a/a
  +++ b/a
  @@ -1,1 +1,1 @@
  - a
  +  a
  diff --git a/b b/b
  index 78981922613b2afb6025042ff6bd878ac1994e85..f76dd238ade08917e6712764a16a22005a50573d
  GIT binary patch
  literal 1
  Ic${MZ000310RR91
  
  diff --git a/c b/c
  --- a/c
  +++ b/c
  @@ -1,1 +1,2 @@
   a
  +

  $ cd ..