I am trying to understand the idea behind Kafka Stream Checkpointing. Is it just a consumer side managing of offset info? Or there is more to it. Any links explaining the details would be helpful.
1 Answers
Checkpoint in Kafka Streams is used for storing offset of changelog topic of state store, so when application restarted and state restore is happened, a restore consumer will try to continue consume from this offset stored in checkpoint file if the offset is still valid, if not the restore process will remove the old state and start restore by consuming from the beginning of changelog topic.
I can't find anything about checkpoint in Kafka Streams document, it was added from 1.1.0.
Related