How to use cpu pinning for burstable or best effort pods, to improve the performance of the pod on openshift?
I did cpu manager enabled for cpu pinning on openshift 4.12.46 , now that works only for guranteed pods, and does not works for bursatble pods, how do we fix this to improve performance and latency of the LLM model?
Good performance for guaranteed pods Expected : Performance improvement for burstable pods as well