search

gluster / project-infrastructure

Issues related to GlusterFs infrastructure components.
0 stars 0 forks source link

Upgrade RHEL 7 CI to latest RHEL/CentOS 7 #153

Open mykaul opened 2 years ago

mykaul commented 2 years ago

Just wanted to make sure we are using 7.9, with latest updates.

Specifically, unsure we are using latest JDK (for Jenkins), OpenSSL (which may explain some crashes we see? Unsure), kernel, etc.

mscherer commented 2 years ago

it should be automatically updated by cron. However, we do not reboot servers unless needed, and only the newer builder restart the agent on demand (a trick we use to cope with openjdk upgrade).

We can do a pass and see if they are all up to date, but more informations could help to narrow down the problems.

mykaul commented 2 years ago

A monthly update + reboot makes sense to me.

mscherer commented 2 years ago

Updates are covered. Reboot, I guess we need to find how to not interfere with running jobs.

If we could schedule 1 job per node on a regular basis instead of 1 node per label, that would be perfect. And ideally, 1 transient job. Maybe using a groovy script.

mscherer commented 2 years ago

https://www.jenkins.io/doc/pipeline/examples/#trigger-job-on-all-nodes seems to be doable. @deepshikhaaa , since you looked at pipeline, do you think that would be doable ? (eg, add a reboot job, and add some pipeline to reboot the nodes ?)

deepshikhaaa commented 2 years ago

Yeah, that's doable! I can work on it next week.