On Tue, Oct 15, 2024 at 04:47:59PM +0200, Peter Krempa wrote:
On Tue, Oct 15, 2024 at 13:09:57 +0100, Daniel P. Berrangé wrote:
> On Tue, Oct 15, 2024 at 01:58:00PM +0200, Peter Krempa wrote:
> > On Tue, Oct 15, 2024 at 12:19:24 +0100, Daniel P. Berrangé wrote:
[...]
> > >
> > > "Only ghcr.io/cirruslabs/macos-runner:sonoma is allowed.
> > > Automatically upgraded."
> > >
> > > Replace (failed) testing of multiple macOS versions, with testing
> > > of multiple XCode versions - 15.4 and 16.0
> >
> > So are these so much different? I'm asking because I'm a bit annoyed
of
> > the CI minutes always running out towards the end of the month.
>
> Basically corresponds to different clang versions.
>
> If we're reliably running out of CI minutes, then dropping down to
> just 1 job is our only way out which doesn't involve paying $$$ for
> more CI time.
I've went through the pipeline logs and we've run out of CI minutes on
cirrus on following dates:
2024/09/23
2024/08/27
2024/07/30
2024/06/27
2024/05/28
It was okay in April and I didn't check earlier, but most likely this
was the start of it, because in May I've also pushed patch decoupling
the 'pages' job from the rest of the CI as our web didn't update due to
CI minutes being out on Cirrus.
It resets at the beginning of the month, so we're marginally over every
month, so shedding one job might help.
The macOS jobs are massively more expensive on credits than FreeBSD
jobs too, so definitely the most benefit is from shedding a macOS
job.
With regards,
Daniel
--
|:
https://berrange.com -o-
https://www.flickr.com/photos/dberrange :|
|:
https://libvirt.org -o-
https://fstop138.berrange.com :|
|:
https://entangle-photo.org -o-
https://www.instagram.com/dberrange :|