-
Bug
-
Resolution: Unresolved
-
Normal
-
None
-
4.10.z
-
Quality / Stability / Reliability
-
False
-
-
3
-
Moderate
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Description of problem:
After restart of one of a masters hosts metal3 pod is restarted as well. I see that new schema is created. None of old are not removed. The new schema is not assigned to any host's HFS
$ oc get pods -n openshift-machine-api -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
cluster-autoscaler-operator-76f76d6747-mdhw4 2/2 Running 0 15h 10.129.0.9 openshift-master-0 <none> <none>
cluster-baremetal-operator-c4bc97f9-7zvsg 2/2 Running 0 15h 10.129.0.5 openshift-master-0 <none> <none>
machine-api-controllers-64f578fff7-xq8fk 7/7 Running 0 15h 10.128.0.10 openshift-master-2 <none> <none>
machine-api-operator-d676654f5-ff6jk 2/2 Running 0 15h 10.129.0.7 openshift-master-0 <none> <none>
metal3-b56d755df-wwzmq 9/9 Running 0 30m 10.46.29.130 openshift-master-1 <none> <none>
metal3-image-cache-hbw58 1/1 Running 0 15h 10.46.29.131 openshift-master-2 <none> <none>
metal3-image-cache-mz4sw 1/1 Running 1 15h 10.46.29.130 openshift-master-1 <none> <none>
metal3-image-cache-p8nr9 1/1 Running 0 15h 10.46.29.129 openshift-master-0 <none> <none>
metal3-image-customization-666c6584bc-95bs9 1/1 Running 0 15h 10.129.0.40 openshift-master-0 <none> <none>
$ oc get hfs -n openshift-machine-api
NAME AGE
openshift-master-0 15h
openshift-master-1 15h
openshift-master-2 15h
openshift-worker-0 15h
openshift-worker-1 15h
$oc get firmwareschema -n openshift-machine-api
NAME AGE
schema-49c0740e 15h
schema-4f53cda1 61m
schema-5179a96a 15h
$ oc describe hfs openshift-master-1 -n openshift-machine-api |grep schem
f:schema:
Name: schema-49c0740e
$ oc describe hfs openshift-master-0 -n openshift-machine-api |grep schem
f:schema:
Name: schema-49c0740e
$ oc describe hfs openshift-master-2 -n openshift-machine-api |grep schem
f:schema:
Name: schema-49c0740e
$ oc describe hfs openshift-worker-0 -n openshift-machine-api |grep schem
f:schema:
Name: schema-49c0740e
$ oc describe hfs openshift-worker-1 -n openshift-machine-api |grep schem
f:schema:
Name: schema-5179a96a
Version-Release number of selected component (if applicable):
4.10.0-0.nightly-2022-11-10-124234
How reproducible:
~25%
Steps to Reproduce:
1. Deploy a cluster. Check how many schemas objects are created 2. Restart one of masters. 3. Check the number of schemas after both master and metal3 pod are back to Running state
Actual results:
New not assigned to any host's HFS schema created
Expected results:
No change in schemas number OR the new schema should replace one of old schemas
Additional info:
1. test run on real BM setup (HP) 2. Will add must-gather 3. Going to re-run the same test to check how often it happens 4. Thinking about testing in other version 5. Waited for about an hour in hope it will be deleted by some kind of garbage collection