Hi all SIGs!
As we all switched (for SIGs) to gitlab.com in the last months (and
final archive was done for all /rpms/* and other projects still on
pagure/git.centos.org) some SIGs started to explore other gitlab
features, like CI/CD.
We got this week a warning from Gitlab about our CentOS project almost
reaching the allocated quota in our sponsored account there, but it
should be reset on monthly basis and it shouldn't be blocking any of
your CI/CD pipelines for this week.
But we still need to verify about which SIGs are (ab)using the "compute
minutes" allocated for our account and find a better/optimized way to
let you still use CI/CD features, but maybe by offloading to some
dedicated runners we'd spin up in our own infra .
I just created yesterday ticket
https://gitlab.com/CentOS/infra/tracker/-/issues/1743 to track this (but
need some investigation and time which I had not planned to spent on
gitlab for now)
Actually the top 3 projects/groups under CentOS org on gitlab using
computes minutes for August are :
- https://gitlab.com/CentOS/automotive/src/automotive-image-builder
(~17200 minutes so far)
- https://gitlab.com/CentOS/Hyperscale/rpms/systemd (~5400 minutes so far)
- https://gitlab.com/CentOS/kmods/rpms/kernel (~1300 minutes so far)
We clearly see an exponential growth for compute minutes usage, but we'd
also remind each SIG/project hosted there to be "good citizen" and so
think that they are quotas in place that would eventually impact other
hosted projects.
--
Fabian Arrotin
The CentOS Project | https://www.centos.org
gpg key: 17F3B7A1 | @arrfab[@fosstodon.org]
Due to hardware consolidation (and reuse to prepare upcoming DC move),
we'll have to migrate some VMs to different hypervisors.
Publicly accessible VMs that SIGs are interacting with and that are
impacted by that operation (other internal VMs will be also migrated but
no impact for SIGs) :
- https://cbs.centos.org
- https://lists.centos.org
Migration is scheduled for """"Tuesday September 2nd, 6:00 am UTC time"""".
You can convert to local time with $(date -d '2025-09-02 06:00 UTC')
The expected "downtime" is estimated to ~30 minutes , time to shutdown
VM/service, migrate definition and storage to different hypervisor,
redefine VM and restart it.
Thanks for your understanding and patience.
on behalf of the Infra team,
--
Fabian Arrotin
The CentOS Project | https://www.centos.org
gpg key: 17F3B7A1 | @arrfab[@fosstodon.org]