Discussed here: https://suse.slack.com/archives/C079ANFDS2C/p1785305851855819
- Document the resource requirements:
- separate per node agents (daemonsets) and cluster-wide agents (deployments, statefulsets)
- Documentation should use a similar table as we use for the platform/
- Check the resource requirements in the same way as we do for the platform (it’s part of the helm chart workflows):
- Total (over all agent pods) average requested resources (not the limit) per default node should be less than 3% of a default node (cpu, memory)
- If we are at more than 3% for cpu OR memory first discuss
- Mark Bakker needs to review as well
STAC-25481
Discussed here: https://suse.slack.com/archives/C079ANFDS2C/p1785305851855819