Re: Fedora-Atomic 27-20180407.0 compose check report

[Adam Williamson]

On Sat, 2018-04-07 at 09:00 -0700, Adam Williamson wrote:
> On Sat, 2018-04-07 at 10:32 -0400, Dusty Mabe wrote:
> > On 04/07/2018 03:54 AM, Fedora compose checker wrote:
> > > No missing expected images.
> > >
> > > Passed openQA tests: 2/2 (x86_64)
> >
> > Can we get these emails to go to [e-mail]?
>
> They're actually supposed to, and they're also only supposed to go out
> when a test fails. I'm not sure whey they're acting like the
> Rawhide/Branched mails at the moment, I had a brief look at it the
> other day but couldn't see anything wrong. I'll have another look at it
> next week.

OK, it wasn't quite 'next week', but :P

I figured out why this was broken, finally. The 'special' configuration
for the two-week Atomic reports (mail them to a different set of
addresses, and only mail on failure) relied on the fedfind 'milestone'
for these composes being 'Atomic', but I changed that back in
https://pagure.io/fedora-qa/fedfind/c/8c979e22cba0138b0bdc8414c86048a4a17f1b...
and
https://pagure.io/fedora-qa/fedfind/c/26c556019bd9c4e4b9fd8cc05164bf18630c1c...
, all for terribly good reasons of course. I've now tweaked check-
compose so you can override the configuration by the release's "dist"
(which is what Pungi/productmd refer to as the 'shortname' - it's the
first component of the compose ID, so 'Fedora', or 'Fedora-Atomic', or
'Fedora-Docker', etc.) instead of its "milestone", and tweaked the
configuration of the check-compose deployment in Fedora infra to match.
So if I got all that right, this should now work again.

I also updated the list of addresses the mails will be sent to; it was
previously Mike McGrath plus atomic@ , but I don't think Mike's the
right person any more, so I changed it to Dusty, Colin and atomic@. So
Dusty, Colin and the list should get the reports *only if* either of
the tests failed. When both tests pass, no report will be sent out.

Thanks folks!

Fedora-Atomic 27-20180407.0 compose check report

[Dusty Mabe]

On 04/25/2018 05:17 PM, Adam Williamson wrote:
> On Sat, 2018-04-07 at 09:00 -0700, Adam Williamson wrote:
> OK, it wasn't quite 'next week', but :P
>
> I figured out why this was broken, finally. The 'special' configuration
> for the two-week Atomic reports (mail them to a different set of
> addresses, and only mail on failure) relied on the fedfind 'milestone'
> for these composes being 'Atomic', but I changed that back in
> https://pagure.io/fedora-qa/fedfind/c/8c979e22cba0138b0bdc8414c86048a4a17f1b...
> and
> https://pagure.io/fedora-qa/fedfind/c/26c556019bd9c4e4b9fd8cc05164bf18630c1c...
> , all for terribly good reasons of course. I've now tweaked check-
> compose so you can override the configuration by the release's "dist"
> (which is what Pungi/productmd refer to as the 'shortname' - it's the
> first component of the compose ID, so 'Fedora', or 'Fedora-Atomic', or
> 'Fedora-Docker', etc.) instead of its "milestone", and tweaked the
> configuration of the check-compose deployment in Fedora infra to match.
> So if I got all that right, this should now work again.
>
> I also updated the list of addresses the mails will be sent to; it was
> previously Mike McGrath plus atomic@ , but I don't think Mike's the
> right person any more, so I changed it to Dusty, Colin and atomic@. So
> Dusty, Colin and the list should get the reports *only if* either of
> the tests failed. When both tests pass, no report will be sent out.
>
> Thanks folks!

Thanks Adam