Environement:

• OCP v4.6.8
• openshift-storage (Ceph)

A cluster that is now not running application workload and only the OCP environment itself is producing a relatively big amount of infra logs and the storage is almost full. Storage is not released after indices are removed manually by a REST API call for the ES.

We can see some TLS connection errors in the logging mast-gather for the curator.
Storage size: 200G, SingleRedundancy Policy, retentionPolicy 1d for infra.

The must-gather and logging mast-gather are available on http://file.brq.redhat.com/rludva/02899863/