Stéphane Derosiaux created KAFKA-9639:
-----------------------------------------
Summary: Task is not killed despite being dead
Key: KAFKA-9639
URL: https://issues.apache.org/jira/browse/KAFKA-9639
Project: Kafka
Issue Type: Bug
Components: KafkaConnect
Affects Versions: 2.3.1
Reporter: Stéphane Derosiaux
After restarting my postgresql, my Debezium connector failed and stopped (no
more metrics, and see the logs below), but the task and the connector in
Connect were still "RUNNING".
It seems the task "stop" procedure was interrupted by the Kafka Connect
framework and let the connector in a "zombie" state.
{code:java}
WorkerSourceTask\{id=customer-order-cdc-0} Task threw an uncaught and
unrecoverable exception (org.apache.kafka.connect.runtime.WorkerTask:179)
org.apache.kafka.connect.errors.ConnectException: An exception occurred in the
change event producer. This connector will be stopped.
Caused by: org.postgresql.util.PSQLException: Database connection failed when
reading from copy
WorkerSourceTask\{id=customer-order-cdc-0} Task is being killed and will not
recover until manually restarted
(org.apache.kafka.connect.runtime.WorkerTask:180)
Creating thread debezium-postgresconnector-customer-order-dev-error-handler
(io.debezium.util.Threads:268)
Connector has already been stopped
(io.debezium.connector.postgresql.PostgresConnectorTask:241)
Finished streaming (io.debezium.pipeline.ChangeEventSourceCoordinator:100)
Connected metrics set to 'false'
(io.debezium.pipeline.metrics.StreamingChangeEventSourceMetrics:59)
[2020-03-02 08:48:08,074] ERROR [customer-order-cdc|task-0] Interrupted while
stopping (io.debezium.connector.postgresql.PostgresConnectorTask:263)
java.lang.InterruptedException
[Producer clientId=connector-producer-customer-order-cdc-0] Closing the Kafka
producer with timeoutMillis = 30000 ms.
(org.apache.kafka.clients.producer.KafkaProducer:1153)
{code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
