Hello Maximo folks,

Wanted to check if anybody has customized the cpu resource and limits of pods required for MAS. By default, all pods created in installation process considers the alot of additional buffer which at present we see is not required. Because of that our pods are not able to schedule. Although the cpu utlization is very less but still we get insufficient cpu error message in pod as other pods have requested alot of additional buffer.

If anybody has customized these CPU resource and limit settings, please share your inputs/suggestions or any links we can refer.

Hi Poonam,

I experimented with pod templates during our initial R&D for the upgrade but didn't achieve much success. Each improvement really just seemed to create additional complications. Here is some documentation on the custom workloads and pod templates for the CRDs that deploy core and manage services if you'd like to review it. Ultimately, I believe we resolved our issues by just adding more compute resources.