-
Bug
-
Resolution: Done
-
Major
-
None
-
None
-
None
-
None
-
False
-
None
-
False
-
0
-
0
-
0
Memory of some HCP pods are undersized and they consume more than 100% of requested memory size on creation itself,
- hosted-cluster-config-operator = ~600%
- certified-operators-catalog = ~340%
- packageserver = ~220%
- cluster-autoscaler = ~200%
- cluster-network-operator = ~200%
Steps to reproduce,
- Deploy a 3 node(m5.4xl) management cluster on AWS(4.10.0-0.nightly-2022-06-01-204608)
- Create a HostedCluster using OVN(4.11.0-0.nightly-2022-05-25-123329) with nodepool size of 7
- Watch HCP pod usage for few hours
Snapshot captured on AWS hostedcluster - https://snapshots-origin.grafana.net/dashboard/snapshot/TEEBeler7fvKyIp8Jy4DOrgkuuOgTaFe?orgId=2