My source connector throws
Caused by: io.confluent.kafka.schemaregistry.client.rest.exceptions.RestClientException: Error while forwarding register schema request to the master; error code: 50003
or
Caused by: io.confluent.kafka.schemaregistry.client.rest.exceptions.RestClientException: Master not known
I found this happened when schema registry's master changes and I have two replicas of schema-registry, under the same service on k8s.
The top exception is org.apache.kafka.connect.errors.ConnectException: Tolerance exceeded in error handler
How to increase the tolerance so the connector can retry more times until the new master is elected?