>> Any reason not to draw a line in the sand for the next stabilization release (3.9)?
+1 Josh, add me to the "test fixers" queue, as well. However, I think the authors of patches that break the build should also be on the hook for fixing problems, as well. On Wed, Jul 20, 2016 at 2:05 PM, Jonathan Ellis <jbel...@gmail.com> wrote: > Also +1 from me. Any reason not to draw a line in the sand for the next > stabilization release (3.9)? > > On Wed, Jul 20, 2016 at 3:11 PM, sankalp kohli <kohlisank...@gmail.com> > wrote: > > > Hi, > > I am +1 for not releasing new version till we fix the unit tests. > Yes I > > am interested in being added to the queue to make testing better. I have > > fixed a bunch of flaky tests in our internal build and will open source > > them soon. > > > > Can the PMC give some guidance on making releases block on unit tests? > > > > Also I see many unit tests commented out by 8099 like > StreamTransferTest. I > > dont know the history there so want to know if this is intentional? > > > > Thanks, > > Sankalp > > > > On Wed, Jul 20, 2016 at 8:35 AM, Josh McKenzie <jmcken...@apache.org> > > wrote: > > > > > I've recently started a hard push for getting test tickets prioritized > > and > > > assigned to various contributors, and I'm pushing for us to have a > > > consistent green test board for 3.9. While I'm not on the PMC so can't > > > binding -1 a release vote, I'd non-binding -1 3.9 if we don't have > stable > > > tests. > > > > > > After that release my focus is going to shift to trunk (3.10) for > > > stabilization; it's definitely something that's become a higher > priority > > > within the past month or so. > > > > > > If you or anyone else on the dev list here would like to be included in > > the > > > test-fixing assignment queue to help speed up the process (and get to > > know > > > parts of the code-base you don't yet know!), please let me know and > I'll > > > start kicking out test fix tickets your way once triaged. > > > > > > Thanks. > > > > > > On Tue, Jul 19, 2016 at 7:08 PM, sankalp kohli <kohlisank...@gmail.com > > > > > wrote: > > > > > > > Hi All, > > > > I can see that unit tests are very flaky in trunk as well > as > > > > previous versions of Cassandra. Out of the last 100 builds(#949 to > > #1049) > > > > for trunk, I only saw 1 successful build. > > > > > > > > Is there any plan to fix the failing tests that I am not aware of? I > > see > > > > lots of tickets recently for failing dtests, and am hoping the unit > > tests > > > > can receive similar attention. > > > > > > > > Thanks, > > > > Sankalp > > > > > > > > > > > > > -- > Jonathan Ellis > Project Chair, Apache Cassandra > co-founder, http://www.datastax.com > @spyced >
