Hi, in my flink job checkpointing is enabled and i have set this two configs seperatly •

commit.offsets.on.checkpoint: false

and

enable.auto.commit: true

and

<http://auto.commit.interval.ms|auto.commit.interval.ms>: 10

. •

commit.offsets.on.checkpoint: true

now with this configurations in both cases data is being duplicated in the sink topic during failure recovery, so is there any way to avoid duplidate data?

Flink doesn't rely on committing offsets for its fault tolerance. If you want to avoid duplicate data, you can consider enabling exactly once as your processing guarantee