Re: Killing obsolete Jenkins builds

On Thu, Nov 19, 2015 at 5:39 AM, Stephan Bergmann <sbergman@redhat.com>
wrote:

> By the way, one situation where it is debatable whether all the triggered
> builds are useful is if you push a series of changes to gerrit, and Jenkins
> does builds for each of the changes in the series. For me at least, in such
> a situation it would suffice if Jenkins just only did a build for the
> topmost change.

Yes, good point. This could be helped by the plugin Samuel mentioned. But
it'd only work when you know what you're doing and when you are a core
developer with commit rights (I'd expect patches submitted by the community
to pass builds/tests before getting reviewed in most cases).


On Thu, Nov 19, 2015 at 4:39 AM, Stephan Bergmann <sbergman@redhat.com>
 wrote:

> On 11/18/2015 03:28 PM, Ashod Nakashian wrote:
>
> > As an aside, the recent drop in average build times (visible from the
> > Jenkins build times list/graph) is something I like to believe is my
> > fault :) The relevant commits were pushed on Sunday last and now there
> > are far more sub 1h builds and the peaks for longer builds are shorter.
> > I like to reduce that average even further by avoiding wasting the
> > machine cycles on irrelevant work.
>
> As far as I understand, Jenkins builds for gerrit changes are not complete
> rebuilds, but are done as incremental builds on top of what previous builds
> (of typically unrelated gerrit changes) happened to leave behind.  That
> likely makes analysing such build times hard.


Interesting. Hadn't noticed. It seems Jenkins isn't doing a great job here.
At any rate, the way I did my unscientific analysis is by looking at the
high and low plateaus. The low was around 30 minutes, and this didn't
change. The high was around 70-80 minutes and now it's closer to 60
minutes. But I agree it's not easy to analyze. But I would hope people
noticed their build times went down.