Closed techietav closed 8 months ago
I have confirmed that installing knative eventing resolves this problem.
I managed to get WA installed by following this. https://www.ibm.com/docs/en/cloud-paks/cp-data/4.8.x?topic=software-installing-red-hat-openshift-serverless-knative-eventing
cpd-cli manage authorize-instance-topology \ --release=${VERSION} \ --cpd_operator_ns=ibm-knative-events \ --cpd_instance_ns=knative-eventing
cpd-cli manage setup-instance-topology \ --release=${VERSION} \ --cpd_operator_ns=ibm-knative-events \ --cpd_instance_ns=knative-eventing \ --license_acceptance=true
cpd-cli manage deploy-knative-eventing \ --release=${VERSION} \ --block_storage_class=${STG_CLASS_BLOCK}
I also discovered that I in ibm-knative-events namespace > IBM NamespaceScope Operator I needed to update the value of MultiNamespace to true as it was in error.
spec
installModes:
type: MultiNamespace
- supported: true
@techietav Can you clarify which CR you updated in the ibm-knative-events
namespace? Is it the NameSpaceScope?
I see that you're mixing CP4D instance and operator namespaces with the Knative eventing namespaces. I would not recommend that. Suggest to use cpd
for the CP4D instance namespace and cpd-operators
for the CP4D operators namespace. The other knative
namespaces are created by the deploy-knative-eventing
script.
@techietav My installation finished ok without having to make any further changes.
The CP4D documentation says to use those namespaces so I followed that.
I see: https://www.ibm.com/docs/en/cloud-paks/cp-data/4.8.x?topic=software-installing-red-hat-openshift-serverless-knative-eventing. I didn't run these steps explicitly on a 4.8.1 installation and found that he OperatorGroup was created correctly:
apiVersion: operators.coreos.com/v1
kind: OperatorGroup
metadata:
annotations:
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"operators.coreos.com/v1","kind":"OperatorGroup","metadata":{"annotations":{},"name":"common-service","namespace":"ibm-knative-events"},"spec":{"targetNamespaces":["ibm-knative-events","knative-eventing"],"upgradeStrategy":"Default"}}
olm.providedAPIs: Kafka.v1beta1.ibmevents.ibm.com,KafkaBridge.v1alpha1.ibmevents.ibm.com,KafkaBridge.v1beta2.ibmevents.ibm.com,KafkaConnect.v1beta1.ibmevents.ibm.com,KafkaConnector.v1alpha1.ibmevents.ibm.com,KafkaMirrorMaker.v1beta1.ibmevents.ibm.com,KafkaMirrorMaker2.v1alpha1.ibmevents.ibm.com,KafkaRebalance.v1alpha1.ibmevents.ibm.com,KafkaTopic.v1beta1.ibmevents.ibm.com,KafkaUser.v1beta1.ibmevents.ibm.com
creationTimestamp: "2024-01-26T07:12:04Z"
generation: 1
name: common-service
namespace: ibm-knative-events
resourceVersion: "38959423"
uid: efd455c5-d429-4b89-ac95-2660dda86f02
spec:
targetNamespaces:
- ibm-knative-events
- knative-eventing
upgradeStrategy: Default
status:
lastUpdated: "2024-01-26T07:12:04Z"
namespaces:
- knative-eventing
- ibm-knative-events
It looks like the steps are redundant.
I have confirmed that installing knative eventing resolves this problem.
I managed to get WA installed by following this. https://www.ibm.com/docs/en/cloud-paks/cp-data/4.8.x?topic=software-installing-red-hat-openshift-serverless-knative-eventing
cpd-cli manage authorize-instance-topology --release=${VERSION} --cpd_operator_ns=ibm-knative-events --cpd_instance_ns=knative-eventing
cpd-cli manage setup-instance-topology --release=${VERSION} --cpd_operator_ns=ibm-knative-events --cpd_instance_ns=knative-eventing --license_acceptance=true
cpd-cli manage deploy-knative-eventing --release=${VERSION} --block_storage_class=${STG_CLASS_BLOCK}
I had to do these steps on the cp4d 4.8.0 as the whole knative stuff was missing. I didnt test the newer versions yet
Describe the bug Installing watson assistant cartridge fails on 4.8.x as it is missing the require Knative operator
https://www.ibm.com/docs/en/cloud-paks/cp-data/4.8.x?topic=manage-deploy-knative-eventing If you plan to install the IBM® watsonx Assistant service on IBM Cloud Pak® for Data, you must install the following software on your cluster to set up a message broker for IBM watsonx Assistant.
During deployment without the pre reqs installed the clu_training component fails to deploy and is stuck in a cycle.
To Reproduce Set watson assistant to installed in the deployer yaml