Description of problem:

The ceph-csi-controller-manager pod keep OOMKilled 

The OCP platform infrastructure and deployment type

Cluster is Bare Metal, installed with Assited Installer, and upagraded since 4.11

The ODF deployment type (Internal, External, Internal-Attached (LSO), Multicluster, DR, Provider, etc):

Internal ODF using 3 worker nodes and NVMe devices

The version of all relevant components (OCP, ODF, RHCS, ACM whichever is applicable):

cephcsi-operator.v4.17.0-rhodf

Does this issue impact your ability to continue to work with the product?

I don't know 

Is there any workaround available to the best of your knowledge?

We tried to bump the limits for the pod, based on https://access.redhat.com/solutions/7002548, but even with x30 of the orignal values, it still fails.

Can this issue be reproduced? If so, please provide the hit rate

We don't have another cluster to verify it. We were hit by this on our production cluster. 

Can this issue be reproduced from the UI?

If this is a regression, please provide more details to justify this:

Steps to Reproduce:

1.

2.

3.

The exact date and time when the issue was observed, including timezone details:

Actual results:

Expected results:

Logs collected and log location:

Additional info: