[openstack-dev] [tripleo][qa] Enabling elastic-recheck for tripleo job

Paul Belanger pabelanger at redhat.com
Tue Aug 8 11:59:30 UTC 2017


On Mon, Aug 07, 2017 at 03:28:13PM -0400, Paul Belanger wrote:
> On Mon, Aug 07, 2017 at 10:34:59AM -0700, Emilien Macchi wrote:
> > On Mon, Aug 7, 2017 at 9:18 AM, Paul Belanger <pabelanger at redhat.com> wrote:
> > > I've just pushed up a change to puppet-elastic_recheck[1] to update the jobs
> > > regex for elastic-recheck to start tracking tripleo.  What this means, is now
> > > the elasticRecheck bot will start leaving comments on reviews once it matches a
> > > failure in knows about.
> > 
> > Really cool, thanks.
> > Note we have an etherpad to track old queries in logstash:
> > https://etherpad.openstack.org/p/tripleo-ci-logstash-queries
> > I know it's not the best way to handle it and we should use more
> > elastic-recheck, but we can also re-use some queries if needed and
> > remove this etherpad, with this move to elastic-recheck.
> 
> Yes, once we have the current elastic-recheck bugs commenting on tripleo
> reviews, we should migrated any existing logstash queries into it.
> 
Just a heads up, this is now live. You'll start to see comments from 'Elastic
Recheck' bot on failed jobs. Please take a moment to look into the failures that
elastic-recheck is reporting, and if the failures is 'unrecognized error',
please dive a little deeper before typing 'recheck'.

The idea now is the start classifying failures, in an attempt to find patterns.
Chances are, if you are hitting an 'unrecognized error' somebody else is hitting
it too.  And what issues have been classified, we can then triage the issue and
see what is needed to fix it.



More information about the OpenStack-dev mailing list