Bug 6083 - `debugwaitonrepack` can returns too early
Summary: `debugwaitonrepack` can returns too early
Status: RESOLVED FIXED
Alias: None
Product: Mercurial
Classification: Unclassified
Component: Mercurial (show other bugs)
Version: 4.9rc0
Hardware: PC Linux
: normal bug
Assignee: Bugzilla
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2019-02-18 05:00 UTC by Boris Feld
Modified: 2020-02-21 00:01 UTC (History)
3 users (show)

See Also:
Python Version: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Boris Feld 2019-02-18 05:00 UTC
While trying to stabilize output of the `test-remotefilelog-bgprefetch.t` test (https://phab.mercurial-scm.org/D5960), we detected that `debugwaitonrepack` is buggy.

Here is how I reproduced it:

    python run-tests.py --runs-per-test=100 test-remotefilelog-bgprefetch.t

I cannot reproduce the issue 100% of the time and it seems more flaky with a busy workstation.
Comment 1 Boris Feld 2019-03-12 05:18 UTC
My patch to try to stabilize the test file has landed and the test is more stable but is still failing from time to time (https://ci.octobus.net/job/MercurialPy2/327/console for example).

From a discussion with Augie, it might be because of using exotic file-system (inside Docker) and maybe linked to flock usage.
Comment 2 Bugzilla 2019-04-12 00:01 UTC
Bug was set to UNCONFIRMED for 30 days, bumping
Comment 3 Bugzilla 2019-05-13 00:01 UTC
Bug was set to UNCONFIRMED for 31 days, bumping
Comment 4 Bugzilla 2019-06-12 00:00 UTC
Bug was set to UNCONFIRMED for 30 days, bumping
Comment 5 Bugzilla 2019-07-13 00:01 UTC
Bug was set to UNCONFIRMED for 31 days, bumping
Comment 6 Bugzilla 2019-08-12 00:01 UTC
Bug was set to UNCONFIRMED for 30 days, bumping
Comment 7 Boris Feld 2019-08-21 12:56 UTC
The CI job now run on a normal VM running Debian 9.9 on an ext4 partition so it might not be related to the file-system in the end,
Comment 8 Bugzilla 2019-09-21 00:01 UTC
Bug was set to UNCONFIRMED for 30 days, bumping
Comment 9 Pierre-Yves David 2019-10-02 07:56 UTC
We no longer use flock. Boris, should we sloce this bug ?
Comment 10 HG Bot 2020-02-13 14:20 UTC
Fixed by https://mercurial-scm.org/repo/hg/rev/ee0959e7d435
Pierre-Yves David <pierre-yves.david@octobus.net>
remotefilelog-test: glob some flaky output line (issue6083)

The two following lines are flaky underload, yet the final result is correct.
The command involves background pre-check of output, these are not stable
probably because they run in parallel in multiple process.

I spent a couple of hours trying to understand the pattern and gave up. The
documented intend of these tests is safely guaranteed by checking the cache
content after the command.

If it become useful to start testing precise internal details of the, they will
have to be tested in a more appropriate framework than `.t` tests.

Differential Revision: https://phab.mercurial-scm.org/D8102

(please test the fix)
Comment 11 Bugzilla 2020-02-21 00:01 UTC
Bug was set to TESTING for 7 days, resolving