tests/test-remote-hidden.t

hgwebdir: avoid systematic full garbage collection

Forcing a systematic full garbage collection upon each request
can serioulsy harm performance. This is reported as
https://bz.mercurial-scm.org/show_bug.cgi?id=6075

With this change we're performing the full collection according
to a new setting, `experimental.web.full-garbage-collection-rate`.
The default value is 1, which doesn't change the behavior and will
allow us to test on real use cases. If the value is 0, no full garbage
collection occurs.

Regardless of the value of the setting, a partial garbage collection
still occurs upon each request (not attempting to collect objects from
the oldest generation). This should be enough to take care of
reference cycles that have been created by the last request
(assessment of this requires changing the setting, not to be 1).

In my experience chasing memory leaks in Mercurial servers,
the full collection never reclaimed any memory, but this is with
Python 3 and biased towards small repositories.

On the other hand, as explained in the Python developer docs [1],
frequent full collections are very harmful in terms of performance if
lots of objects survive the collection, and hence stay in the
oldest generation. Note that `gc.collect()` is indeed trying to
collect the oldest generation [2]. This happens usually in two cases:

- unwanted lingering objects (i.e., an actual memory leak that
  the GC cannot do anything about). Sadly, we have lots of those
  these days.
- desireable long-term objects, typically in caches (not inner caches
  carried by repositories, which should be collected with them). This
  is a subject of interest for the Heptapod project.

In short, the flat rate that this change still permits is
probably a bad idea in most cases, and the default value can
be tweaked later on (or even be set to 0) according to experiments
in the wild.

The test is inspired from test-hgwebdir-paths.py

[1] https://devguide.python.org/garbage_collector/#collecting-the-oldest-generation
[2] https://docs.python.org/3/library/gc.html#gc.collect

Differential Revision: https://phab.mercurial-scm.org/D11204

========================================================
Test the ability to access a hidden revision on a server
========================================================

#require serve

  $ . $TESTDIR/testlib/obsmarker-common.sh
  $ cat >> $HGRCPATH << EOF
  > [phases]
  > # public changeset are not obsolete
  > publish=false
  > [experimental]
  > evolution=all
  > [ui]
  > logtemplate='{rev}:{node|short} {desc} [{phase}]\n'
  > EOF

Setup a simple repository with some hidden revisions
----------------------------------------------------

Testing the `served.hidden` view

  $ hg init repo-with-hidden
  $ cd repo-with-hidden

  $ echo 0 > a
  $ hg ci -qAm "c_Public"
  $ hg phase --public
  $ echo 1 > a
  $ hg ci -m "c_Amend_Old"
  $ echo 2 > a
  $ hg ci -m "c_Amend_New" --amend
  $ hg up ".^"
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ echo 3 > a
  $ hg ci -m "c_Pruned"
  created new head
  $ hg debugobsolete --record-parents `getid 'desc("c_Pruned")'` -d '0 0'
  1 new obsolescence markers
  obsoleted 1 changesets
  $ hg up ".^"
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ echo 4 > a
  $ hg ci -m "c_Secret" --secret
  created new head
  $ echo 5 > a
  $ hg ci -m "c_Secret_Pruned" --secret
  $ hg debugobsolete --record-parents `getid 'desc("c_Secret_Pruned")'` -d '0 0'
  1 new obsolescence markers
  obsoleted 1 changesets
  $ hg up null
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved

  $ hg log -G -T '{rev}:{node|short} {desc} [{phase}]\n' --hidden
  x  5:8d28cbe335f3 c_Secret_Pruned [secret]
  |
  o  4:1c6afd79eb66 c_Secret [secret]
  |
  | x  3:5d1575e42c25 c_Pruned [draft]
  |/
  | o  2:c33affeb3f6b c_Amend_New [draft]
  |/
  | x  1:be215fbb8c50 c_Amend_Old [draft]
  |/
  o  0:5f354f46e585 c_Public [public]
  
  $ hg debugobsolete
  be215fbb8c5090028b00154c1fe877ad1b376c61 c33affeb3f6b4e9621d1839d6175ddc07708807c 0 (Thu Jan 01 00:00:00 1970 +0000) {'ef1': '9', 'operation': 'amend', 'user': 'test'}
  5d1575e42c25b7f2db75cd4e0b881b1c35158fae 0 {5f354f46e5853535841ec7a128423e991ca4d59b} (Thu Jan 01 00:00:00 1970 +0000) {'user': 'test'}
  8d28cbe335f311bc89332d7bbe8a07889b6914a0 0 {1c6afd79eb6663275bbe30097e162b1c24ced0f0} (Thu Jan 01 00:00:00 1970 +0000) {'user': 'test'}

  $ cd ..

Test the feature
================

Check cache pre-warm
--------------------

  $ ls -1 repo-with-hidden/.hg/cache
  branch2
  branch2-base
  branch2-served
  branch2-served.hidden
  branch2-visible
  rbc-names-v1
  rbc-revs-v1
  tags2
  tags2-visible

Check that the `served.hidden` repoview
---------------------------------------

  $ hg -R repo-with-hidden serve -p $HGPORT -d --pid-file hg.pid --config web.view=served.hidden
  $ cat hg.pid >> $DAEMON_PIDS

changesets in secret and higher phases are not visible through hgweb

  $ hg -R repo-with-hidden log --template "revision:    {rev}\\n" --rev "reverse(not secret())"
  revision:    2
  revision:    0
  $ hg -R repo-with-hidden log --template "revision:    {rev}\\n" --rev "reverse(not secret())" --hidden
  revision:    3
  revision:    2
  revision:    1
  revision:    0
  $ get-with-headers.py localhost:$HGPORT 'log?style=raw' | grep revision:
  revision:    3
  revision:    2
  revision:    1
  revision:    0

  $ killdaemons.py