[Gluster-devel] Regression test failures - Call for Action

Jeff Darcy jdarcy at redhat.com
Mon May 4 20:23:49 UTC 2015


> There has been a spate of regression test failures (due to broken tests
> or race conditions showing up) in the recent past [1] and I am inclined
> to block 3.7.0 GA along with acceptance of patches until we fix *all*
> regression test failures. We seem to have reached a point where this
> seems to be the only way to restore sanity to our regression runs.
> 
> I plan to put this into effect 24 hours from now i.e. around 0700 UTC on
> 05/05. Thoughts?

As a complement to this, I suggest that we stop the Jenkins queue and
make the slaves available to people debugging specific failures.  We'll
probably need some way - e.g. an Etherpad somewhere - to coordinate
access so we don't step all over each other.  Also, one of us should
go through the last however-many failures and determine the relative
frequency of failures caused by each test, so we can prioritize.  Any
other volunteers before I spend hours doing it myself?


More information about the Gluster-devel mailing list