Description of Problem: The message in CSO about manila looks like not necessary Version-Release number of selected component (if applicable): 4.6.0-0.nightly-2020-09-01-070508 How Reproducible: Always Steps to Reproduce: 1. Install an OCP4.6 cluster on OSP with manila service enabled. 2. Check the CSO Actual Results: $ oc get co storage -o json| jq .status.conditions [ { "lastTransitionTime": "2020-09-02T02:05:52Z", "reason": "AsExpected", "status": "False", "type": "Degraded" }, { "lastTransitionTime": "2020-09-02T06:08:24Z", "message": "ManilaOLMOperatorRemovalProgressing: CSI driver installed by OLM is not preset", "reason": "AsExpected", "status": "False", "type": "Progressing" }, { "lastTransitionTime": "2020-09-02T05:44:30Z", "message": "ManilaOLMOperatorRemovalAvailable: CSI driver installed by OLM is not preset", "reason": "AsExpected", "status": "True", "type": "Available" }, { "lastTransitionTime": "2020-09-02T02:06:02Z", "reason": "AsExpected", "status": "True", "type": "Upgradeable" } ] If installed manila csi driver operator from operator hub, these messages are still not changed. $ oc get maniladriver NAME AGE cluster 6m18s Expected Results: If we don't support install manila csi driver operator by operator hub in 4.6, these messages might not be needed.
Hi @Qin; When the operator is installed via OLM in 4.5, and carried over to 4.6 via upgrade, CSO will remove the old operator before installing the new (4.6) one. The initial intent of these messages was to report the status of this removal, but I can see how this can be noisy. I've create
(Apologies, my previous commit went through incomplete) Hi @Qin; When the operator is installed via OLM in 4.5, and carried over to 4.6 via upgrade, CSO will remove the old operator before installing the new (4.6) one. The initial intent of these messages was to report the status of this removal, but I can see how this can be noisy. I've created a PR [1] to remove those messages if the old operator wasn't found. Note that I left the message "CSI driver has been removed from OLM" for when the old OLM-based Manila operator is indeed removed by CSO, but that message will only stay there for a short period of time (until the operator reconciles again). Let me know if you think that message should go away as well. [1] openshift/cluster-storage-operator/pull/82
Cluster version: 4.6.0-0.nightly-2020-09-10-195619 ManilaCSIDriverOperator version: 4.6.0-0.nightly-2020-09-10-195619 $ oc get co storage -ojson | jq .status.conditions [ { "lastTransitionTime": "2020-09-11T01:10:35Z", "reason": "AsExpected", "status": "False", "type": "Degraded" }, { "lastTransitionTime": "2020-09-11T01:24:59Z", "reason": "AsExpected", "status": "False", "type": "Progressing" }, { "lastTransitionTime": "2020-09-11T01:13:28Z", "reason": "AsExpected", "status": "True", "type": "Available" }, { "lastTransitionTime": "2020-09-11T01:10:46Z", "reason": "AsExpected", "status": "True", "type": "Upgradeable" } ] Mark VERIFIED according to the above result.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (OpenShift Container Platform 4.6 GA Images), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2020:4196