-
Bug
-
Resolution: Duplicate
-
Major
-
None
-
4.10.0
-
None
-
-
-
Important
-
No
-
Proposed
-
False
-
Description of problem:
Is it possible to configure stateful set in multi-az ocp cluster with cinder csi volume. The pod was restarted and came up on a different AZ. Is there fix so it schedules the pod on the same AZ as the cinder volume? I see these events. Is there a doc on how to configure statefulset in a multiple az ocp clusters with cinder volumes? Is this issue fixed in newer version of openshift/cinder csi driver? ~~~ 57m Normal Scheduled pod/cs-eck-iat-es-default-3 Successfully assigned cs-eck-iat/cs-eck-iat-es-default-3 to ocp-d2edgt0203-f95xk-cs-eck-iat-az2-8tsw5 56m Warning FailedMount pod/cs-eck-iat-es-default-3 Unable to attach or mount volumes: unmounted volumes=[elasticsearch-data], unattached volumes=[downward-api elastic-internal-secure-settings elastic-internal-unicast-hosts data-ingest-endpoint es-monitoring-c00e53-ca elastic-internal-elasticsearch-config elastic-internal-elasticsearch-bin-local elastic-internal-http-certificates elasticsearch-logs elastic-internal-probe-user elastic-internal-scripts injection-config cs-eck-iat-filebeat-config data-ingest-enrichment cs-eck-iat-metricbeat-config oneagent-share elastic-internal-xpack-file-realm oneagent-bin elastic-internal-transport-certificates kb-monitoring-local-ca elastic-internal-remote-certificate-authorities elasticsearch-data elastic-internal-elasticsearch-config-local elastic-internal-elasticsearch-plugins-local]: timed out waiting for the condition 43m Warning FailedAttachVolume pod/cs-eck-iat-es-default-3 AttachVolume.Attach failed for volume "pvc-ed1689e0-5562-40b1-8b7a-f7f5d5ec3d5a" : rpc error: code = Internal desc = [ControllerPublishVolume] Attach Volume failed with error failed to attach 44caa810-f8b2-4773-b6c9-66636c3da2f8 volume to 216f4027-59c2-4a27-a91b-f6a4f295420b compute: Bad request with: [POST https://openstack.osp2.cloud2.oneadp.com:13774/v2.1/servers/216f4027-59c2-4a27-a91b-f6a4f295420b/os-volume_attachments], error message: {"badRequest": {"code": 400, "message": "Invalid volume: Instance 216f4027-59c2-4a27-a91b-f6a4f295420b and volume 44caa810-f8b2-4773-b6c9-66636c3da2f8 are not in the same availability_zone. Instance is in az2. Volume is in az1"}} ~~~
Additional info:
OCP version - 4.10