-
Notifications
You must be signed in to change notification settings - Fork 22
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Hook race fix #21
Hook race fix #21
Conversation
Made scontrol changes, only thing is that by using NodeName=all the other named nodes (i.e slurmd-[2-9]) will get added to the list in a drain state, but I'm guessing this wouldn't really be an issue in production? |
Actually one thing that I've realised is that this fix introduces a NEW race condition where once the nodes are undrained, jobs could potentially get scheduled onto them before the upgrade starts, I need to move the undraining to a post-upgrade hook |
The pre-upgrade hook will now drain all nodes before checking if jobs are running on them in order to prevent a race conditions in which pending jobs are scheduled onto the nodes after the check if complete. Will RESUME nodes on a successful upgrade