Known issues and limitations for watsonx Assistant

1. Edit the namespace scope operator csv file:

   oc edit csv ibm-namespace-scope-operator.v{$CPD_VERSION} -n ibm-knative-events

2. Set the MultiNamespace parameter to true:

     
     installModes:
     - supported: true
       type: MultiNamespace

3. Delete the old namespace-scope-operator pod.

   
   nss_op_pod=$(oc get pods -n ibm-knative-events -l name=ibm-namespace-scope-operator --no-headers | awk '{print $1}')
   oc delete pod $nss_op_pod

4. Rerun the following command:

   cpd-cli manage setup-instance-topology --release=${VERSION} --cpd_operator_ns=ibm-knative-events --cpd_instance_ns=knative-eventing --license_acceptance=true

1. Exec into the docker pod running olm utils.

   docker exec -it olm-utils-play-v3 bash

2. Check for the line that is to be removed.

   cat /opt/ansible/bin/deploy-knative-eventing | grep kafka-broker-dispatcher

   Output:

   cat /opt/ansible/bin/deploy-knative-eventing | grep kafka-broker-dispatcher
   oc wait pods -n knative-eventing --selector app=kafka-broker-dispatcher --for condition=Ready --timeout=60s

3. Remove the line.

   sed -i '/kafka-broker-dispatcher/d' /opt/ansible/bin/deploy-knative-eventing

4. Verify whether the line is removed.

   cat /opt/ansible/bin/deploy-knative-eventing | grep kafka-broker-dispatcher

   Output:

   cat /opt/ansible/bin/deploy-knative-eventing | grep kafka-broker-dispatcher
   

oc patch kafka knative-eventing-kafka -n knative-eventing --type=merge '-p={"spec":{"kafka":{"jvmOptions":{"javaSystemProperties":[{"name": "jdk.nativeDigest", "value": "false"}]}}}}' 

oc patch kafkas.ibmevents.ibm.com knative-eventing-kafka -n knative-eventing --type='merge' -p='{"spec":{"kafka":{"jvmOptions":{"javaSystemProperties":[{"name": "jdk.nativeDigest", "value": "false"}]}}}}' 

1. Check the following paths in the watsonx Assistant CR:

   .spec.cluster.blockStorageClass
   
   .spec.cluster.storageClassName

2. Run the following command to inspect the CR:

   oc get watsonassistants.assistant.watson.ibm.com wa -n cpd -o yaml | grep -A 7 "cluster:"

   If the fields blockStorageClass or storageClassName is missing or set to null, the CR fails the validation.
3. To resolve the issue, pass the appropriate storage class values in the upgrade command by using the following flags:

   cpd-cli manage apply-cr \
   --components=watson_assistant \
   --release=${VERSION} \
   --cpd_instance_ns=${PROJECT_CPD_INST_OPERANDS} \
   --param-file=/tmp/work/install-options.yml \
   --block_storage_class <your-block-storage-class> \
   --file_storage_class <your-file-storage-class>
   --license_acceptance=true \
   --upgrade=true

   These values must match the ones that are used in your current deployment.

1. Run the following command to find the watsonx Assistant Postgres pod in the CrashLoopBackOff state.

   oc get pods --no-headers | grep -Ev "Comp|0/0|1/1|2/2|3/3|4/4|5/5|6/6|7/7|8/8" | grep wa-postgres

   The output looks like this:

   wa-postgres-3 0/1 CrashLoopBackOff 115 (2m30s ago) 9h

2. Run the following command to identify if the Postgres pod is the primary pod:

   oc get cluster | grep wa-postgres

   The output looks like this:

   oc get cluster | grep wa-postgres
   wa-postgres         2d20h   3           3       Cluster in healthy state   wa-postgres-1

   Where wa-postgres-1 is the primary pod.
   Tip: If the primary instance is in CrashLoopBackOff status, do the steps in the Postgres cluster in bad state topic.
3. Delete the non-primary pod and its PersistentVolumeClaim (PVC) to create a new pod that syncs with the primary pod.
   Warning: Do not delete a primary pod because it can lead to database downtime and potential data loss.

   oc delete pod/wa-postgres-3 pvc/wa-postgres-3

   Important: Ensure that the EDB operator is running before deleting the pod and its PVC.