[Gluster-infra] Idea: Failure Trends
nigelb at redhat.com
Wed Aug 17 15:09:31 UTC 2016
On Wed, Aug 17, 2016 at 02:59:15PM +0200, Michael Scherer wrote:
> Le mercredi 17 août 2016 à 06:33 +0530, Nigel Babu a écrit :
> > Hello folks,
> > For the last few weeks, Poornima has been sending in a report of failures over
> > the last week. I've been thinking of how to use automation to make that report
> > better. One of the ideas was to use Jenkins to send the email. It seemed like
> > we were automating the wrong thing there.
> So I likely did skip the mail, but are we speaking of failure in the CI,
> or some others types of failure ?
Only for test failures in the CI.
> > Instead, I propose a small website that'll track failures week-on-week and
> > trends of various failures over time. If a particular failure has increased in
> > number over one week, we'll know something has gone wrong.
> But, if there is failure at all, something is already wrong, or I
> misunderstand the whole idea ?
We have occasional failures that we retry and get green. I'm trying to track
trends of these intermittent failures.
> Michael Scherer
> Sysadmin, Community Infrastructure and Platform, OSAS
More information about the Gluster-infra