• Login to Openshift CLI as administrator.

• export ns=ibm-common-services
  oc delete csv -l operators.coreos.com/ibm-events-operator.$ns='' -n $ns

   
•  ODLM will automatically reconcile and create ibm-events-operator in ibm-common-services namespace.  

• Login to Openshift CLI as administrator.

• export ns=ibm-common-services
  oc get pods | grep -i operand-deployment-lifycle-manager
  oc delete pod <podname> #Replace <podname> from the above output.

   
•  ODLM will automatically reconcile and create operandrequest-kafka-iaf-system and ibm-events-operator subscription and CSV in ibm-common-services namespace.  

• Login to Openshift CLI as administrator.

• export ns=ibm-common-services
  oc get pods | grep -i postgresql-operator-controller-manager
  oc delete pod <podname> #Replace <podname> from the above output.

   
•  Zen operator will automatically reconcile and create the requires labels, annotations for PostgreSQL database for Zen. 

• Update EDB Cluster configuration for zen-metastore-edb under resources limits with an entry of hugepages type along with size value. (For example: hugepages-2Mi: "2Gi"
•  Zen operator will automatically reconciles to complete deployment.

Default self-signed certificates for CPD route will not be renewed automatically when expired

•  oc patch ZenService iaf-zen-cpdservice --type='json' -p='[\{"op": "add", "path": "/spec/zenCustomRoute","value":{"route_secret":"dummy-value","route_reencrypt":true}}]' -n ${CP4BA_AUTO_NAMESPACE}
•   oc patch ZenService iaf-zen-cpdservice --type='json' -p='[\{"op": "add", "path": "/spec/zenCustomRoute","value":{"route_secret":"iaf-system-automationui-aui-zen-cert","route_reencrypt":true}}]' -n ${CP4BA_AUTO_NAMESPACE}

• Login to Openshift CLI as administrator.

• export operatorNamespace=<operator-namespace> # openshift-operators
  export commonuiCSV=$(oc -n $operatorNamespace get csv | grep ibm-commonui-operator | awk '{print $1}')
  oc -n $operatorNamespace patch csv $commonuiCSV --type="json" -p '[{"op": "replace", "path":"/spec/install/spec/deployments/0/spec/template/spec/containers/0/resources/limits/memory", "value":"200Mi"}]'

   
•  common-UI operator pod will start automatically with increased memory.

• When cp4a-prerequisites.sh asks this question (Do you want to use an external Postgres DB [YOU NEED TO CREATE THIS POSTGRESQL DB BY YOURSELF FIRST BEFORE APPLY CP4BA CUSTOM RESOURCE] as BTS metastore DB for this CP4BA deployment?) when you choose to install FNCM (or) ODM with not selecting BAI optional component , please select "No". 
• This is fixed in 24.0.0-IF001

• Go to the <namespace> of the CP4BA deployment that you are upgrading.

  export NAMESPACE = <namespace>
  oc project $NAMESPACE

• Scale down the Business Teams Service (BTS) operator to 0.

  oc scale deployment ibm-bts-operator-controller-manager -n $NAMESPACE --replicas=0

• Delete the Postgres batch job by running the following command.

  oc delete job create-postgres-license-config -n $NAMESPACE

• Delete the ODLM operand pod by running the following command.

  oc delete pod -l name=operand-deployment-lifecycle-manager -n $NAMESPACE

• Check that the Postgres job is restarted by running the following command.

  oc get pod -l job-name=create-postgres-license-config -n $NAMESPACE

• Check the status of the Cloud Pak foundational services databases by running the following command.

  oc get cluster -n $NAMESPACE

• Scale the Business Teams Service (BTS) operator back up to 1.

  oc scale deployment ibm-bts-operator-controller-manager -n $NAMESPACE --replicas=1

Symptom:

After a fresh install or upgrade to CP4BA 24.0.0.x in an All namespaces deployment with multiple CP4BA instances that include BAI,

the BAI instances that are deployed after the first BAI instance may not work as expected,

The error that appears in the “xxx-bai-content-xxx” pod will resemble this:

ERROR    Run request failed.
{
  "message": "Cannot read properties of undefined (reading 'get')",
  "status": 500,
  "name": "UndefinedError",

Cause:

The BAI pods need a restart to read properties in an All namespaces deployment.

Solution:

1. Restart (i.e. delete) “xxx-insights-engine-management-xxx” pod in the namespace of the failing BAI instance
2. Restart (i.e. delete) the “xxx-bai-navigator-xxx” and “xxx-bai-content-xxx” jobs in the namespace of the failing BAI instance

Only perform this workaround for the namespaces where the BAI is not working as expected.

The restart steps can be performed in the OCP console or on the command-line using commands like the following:

(Replace <ns> below with the namespace of the CP4BA instance where BAI is not working as expected)

oc -n <ns> get pods | grep insights-engine-management | awk '{print $1}' | xargs oc -n <ns> delete pod --force --grace-period=0
oc -n <ns> get job | grep bai-navigator | awk '{print $1}' | xargs oc -n <ns> delete job
oc -n <ns> get job | grep bai-content | awk '{print $1}' | xargs oc -n <ns> delete job

• Login to Openshift administration console
• Expand Home -> Search -> Resources to select "OperandRegistry"
• Change the project to CP4BA namespace when upgrading "shared2dedicated" (or) "openshift-operators" when upgrading to "clusterscope2clusterscope"
• Delete the operandRegistry resource "common-service" 
• Expand Workloads -> Pods 
• Delete ibm-common-service-operator pod
• Wait for ibm-common-service-operator pod to be restarted which automatically recreate deleted operandRegistry for ODLM.
• Access generated operandRegistry to check fallbackChannels metadata.
• zenService upgrade will progress with no issues.

constraints not satisfiable: no operators found from catalog ibm-bts-operator-catalog-v3-35-2 in namespace production referenced by subscription ibm-bts-operator, subscription ibm-bts-operator exists

• Login to Openshift administration console
• Expand Home -> Search -> Resources to select "Subscriptions"
• Change the project to CP4BA namespace when upgrading "shared2dedicated"
• Locate PostgreSQL subscription
• Click on Pencil icon for "Update Channel"
• Select the channel -> stable-v1.25
• Wait for the operator to switch the channel and become available. If this process is taking time, restarting Operand Deployment Lifecycle Manager(ODLM) should do the job.
• Wait for the zenService upgrade to finish.