Bug 2292175

+++ This bug was initially created as a clone of Bug #2280834 +++

Description of problem:

After performing 4.14 to 4.16 EUS to EUS upgrade procedure of ODF and OCP, and performing functional level MCG related operations (tier1 tests), noobaa-db-pg-0 got in CrashLoopBackOff


Version of all relevant components (if applicable):

Initial versions installed:
ODF 4.14 + OCP 4.14

Upgraded versions:
OCP 4.16.0-0.nightly-2024-05-15-001800
odf-operator.v4.16.0-101.stable


Is there any workaround available to the best of your knowledge?
Not aware


Can this issue be reproducible?
Tried this procedure with tier1 test execution only once so far


Can this issue be reproduced from the UI?
N/A


If this is a regression, please provide more details to justify this:
Hard to say if it's a regression at this point, as this is the first time we are trying EUS to EUS upgrade 


Steps to Reproduce:
On IBM Cloud VPC:
1. Install OCP 4.14 (IPI), ODF 4.14
2. Upgrade ODF to 4.14, sequentially (4.14 to 4.15, 4.15 to 4.16)
3. Perform OCP EUS to EUS upgrade procedure:

# oc patch clusterversions/version -p '{"spec":{"channel":"stable-4.16"}}' --type=merge

# oc patch mcp/worker --type merge --patch '{"spec":{"paused":true}}'

# oc adm upgrade --to-image=registry.ci.openshift.org/ocp/release:4.16.0-0.nightly-2024-05-15-001800 --allow-explicit-upgrade --force

# oc patch mcp/worker --type merge --patch '{"spec":{"paused":false}}'

Perform functional level operations of NooBaa (tier1 tests of ocs-ci)


Actual results:
As part of the setup phase of test_bucket_creation_deletion.py::TestBucketCreationAndDeletion::test_bucket_creation_deletion[3-S3-DEFAULT-BACKINGSTORE], where the below resources have been created, noobaa-db-pg-0 started crashing.

2024-05-16 14:43:24  tests/functional/object/mcg/test_bucket_creation_deletion.py::TestBucketCreationAndDeletion::test_bucket_creation_deletion[3-S3-DEFAULT-BACKINGSTORE] 
2024-05-16 14:43:24  -------------------------------- live log setup --------------------------------


AWS CLI configMap, s3cli StatefulSet, and AWS, Azure, GCP and IBM Cloud COS secrets.

Started seeing connection to the DB getting broken:

2024-05-16 14:55:52  07:55:40 - ThreadPoolExecutor-30_0 - ocs_ci.utility.utils - INFO  - Executing command: oc --kubeconfig /home/jenkins/current-cluster-dir/openshift-cluster-dir/auth/kubeconfig -n openshift-storage rsh noobaa-db-pg-0 bash -c "pg_dump nbcore | gzip > /tmp/nbcore.gz"
2024-05-16 14:55:52  07:55:41 - ThreadPoolExecutor-30_0 - ocs_ci.utility.utils - WARNING  - Command stderr: error: unable to upgrade connection: container not found ("db")
2024-05-16 14:55:52  
2024-05-16 14:55:52  07:55:41 - ThreadPoolExecutor-30_0 - ocs_ci.ocs.utils - ERROR  - Failed to dump noobaa DB! Error: Error during execution of command: oc --kubeconfig /home/jenkins/current-cluster-dir/openshift-cluster-dir/auth/kubeconfig -n openshift-storage rsh noobaa-db-pg-0 bash -c "pg_dump nbcore | gzip > /tmp/nbcore.gz".
2024-05-16 14:55:52  Error is error: unable to upgrade connection: container not found ("db")



% oc get pod noobaa-db-pg-0     
NAME             READY   STATUS             RESTARTS          AGE
noobaa-db-pg-0   0/1     CrashLoopBackOff   206 (4m58s ago)   17h



% oc describe pod noobaa-db-pg-0

Events:
  Type     Reason   Age                     From     Message
  ----     ------   ----                    ----     -------
  Normal   Pulled   157m (x177 over 17h)    kubelet  Container image "registry.redhat.io/rhel9/postgresql-15@sha256:1aeac23901c0147e4c6e9a1b8bb5f41dd6f95532b0d96adac55d609d0eed32fe" already present on machine
  Warning  BackOff  2m49s (x4742 over 17h)  kubelet  Back-off restarting failed container db in pod noobaa-db-pg-0_openshift-storage(c7d42c55-f3de-45ca-ad7a-408fe7419820)


Noticing this error while trying to get the noobaa-db-pg-0 logs:

Incompatible data directory.  This container image provides
PostgreSQL '15', but data directory is of
version '12'.

This image supports automatic data directory upgrade from
'13', please _carefully_ consult image documentation
about how to use the '$POSTGRESQL_UPGRADE' startup option.



Additional info:
ODF Must Gather - https://url.corp.redhat.com/dc96623
Live cluster details - https://url.corp.redhat.com/331727d

--- Additional comment from Elad on 2024-05-16 14:53:20 UTC ---

Correction:

Steps to Reproduce:
2. Upgrade ODF to **4.16**, sequentially (4.14 to 4.15, 4.15 to 4.16)

--- Additional comment from Danny on 2024-05-20 07:30:16 UTC ---

In the DB logs, it appears that the DB data directory is still in Postgres 12 format and not Postgres 15, a change that should have happened in 4.15. Was the upgrade to 4.15 successful before proceeding to 4.16?