Abhilash Raj pushed to branch master at GNU Mailman / Mailman Core
Commits:
812889d1 by Abhilash Raj at 2019-07-26T06:34:48Z
Fixed a sort-of race condition causing repeated test failures.
This was a very interesting failure, which I think should now be fixed. The
problem seemed to be that since the example plugin is initialized in the
memory and there two worker threads for gunicorn, each of them stored the data
separately with different values.
So, depending on which one served your request, you could get different values
causing test failures when the response was served by a different
worker. Setting the worker threads to 1 during testing should fix this.
- - - - -
f7b9cd03 by Abhilash Raj at 2019-07-26T17:49:34Z
Merge branch 'fix-race' into 'master'
Fixed a sort-of race condition causing repeated test failures.
See merge request mailman/mailman!539
- - - - -
3 changed files:
- src/mailman/plugins/docs/intro.rst
- src/mailman/plugins/testing/rest.cfg
- src/mailman/testing/testing.cfg
Changes:
=====================================
src/mailman/plugins/docs/intro.rst
=====================================
@@ -188,6 +188,7 @@ Plugins and their child resources can support any HTTP
method, such as
number: 7
... or ``DELETE``.
+::
>>> dump_json('http://localhost:9001/3.1/plugins/example/echo',
... method='DELETE')
=====================================
src/mailman/plugins/testing/rest.cfg
=====================================
@@ -4,3 +4,4 @@ enabled: yes
[webservice]
port: 9001
+workers: 1
=====================================
src/mailman/testing/testing.cfg
=====================================
@@ -30,6 +30,7 @@ configuration: python:mailman.testing.passlib
[webservice]
port: 9001
+workers: 1
[runner.archive]
max_restarts: 1
View it on GitLab:
https://gitlab.com/mailman/mailman/compare/278764666922a14fb72ab6fb1b8ac451407041e9...f7b9cd0380130ae1a5ac832e0a4f459c335b7d25
--
View it on GitLab:
https://gitlab.com/mailman/mailman/compare/278764666922a14fb72ab6fb1b8ac451407041e9...f7b9cd0380130ae1a5ac832e0a4f459c335b7d25
You're receiving this email because of your account on gitlab.com.
_______________________________________________
Mailman-checkins mailing list
Mailman-checkins@python.org
Unsubscribe:
https://mail.python.org/mailman/options/mailman-checkins/archive%40jab.org
