After restart of Confluent cluster all previously created and working JDBC Connectors disappeared. My question is: how to prevent such situation in the future? Where are exactly stored configurations of JDBC Connectors? For following topics:
I could use a little more information, but if I’m understanding correctly, when you say disappeared do you mean you can’t find the instances or they just stopped running? I think what may have happened is during the restart the JDBC Connectors shut down. Where is the cluster running? I noticed the name of the topics includes “docker” is this a local deployment?
How did you restart the cluster?
Just guessing, but if you’re using Docker Compose and you did docker-compose down and then docker-compose up that would have removed your Kafka broker on which the connector config is stored in topics.
I forgot to mention earlier: I’ve got Confluent cluster running on Openshift. Kafka Broker has mounted Openshift persistent volume, so I was a little surprised when these topics docker-connect-* were cleared after cluster restart, since I was expecting these data to be persisted.
This topic can be closed since I’ve found reason for this situation. I will write it here just in case someone face similar situation
it turned out that I was not persisting directory on Kafka Broker with a data of all topics. When I was restarting cluster all data on these topics(esp. these topics needed for connectors) was lost and it looked like JDBC Connectors previously declared were lost too.
Managed to solve it by adding persistent volume claim to Openshift deployment config of Kafka broker.