Summary
After redeploying the Transformation Hub capability, the th-schemaregistry-xxx pod is in crash loop backoff in a single node deployment.

Products
ArcSight Transformation Hub

Environment
Arcsight Data Platform 22.1
Transformation Hub 3.6
CDF Version 2021.05.00286

Situation
th-schemaregistry-xxx pod in CrashLoopBackOff status after redeploy of Transformation Hub capability on single node environment.

The following output is seen in the th-schemaregistry-xxx pod logs
[main] INFO io.confluent.admin.utils.ClusterStatus - Expected 2 brokers but found only 1. Trying to query Kafka for metadata again ...
[main] ERROR io.confluent.admin.utils.ClusterStatus - Expected 2 brokers but found only 1. Brokers found [example.net:9093 (id: 1001 rack: null)].

Other Transformation Hub related pods th-kafka-manager-xxx and th-web-service-xxx are in init status.

Event ingestion via the Transformation Hub has stopped.

Resolution
1. Open a command line session on the first master server within the cluster.
Set the variable for the namespace required.
namespace=$( kubectl get namespaces | awk '/^arcsight-installer-/{print $1}' )

2. Edit the th-schemaregistry deployment.

kubectl edit deployments.apps -n $namespace th-schemaregistry

3. Navigate to the required section and change the number of required brokers to a value of "1", like below:

- name: SCHEMA_REGISTRY_CUB_KAFKA_MIN_BROKERS
value: "1"

4. Save and exit the deployment configuration.

5. The schema registry pod will automatically restart, monitor the cluster to observe the status of the schema registry pod and other related pods.

kubectl get pods -n $namespace

Knowledge Base Article Link

URL Name
KM000015474