|Table of Contents|
Current state: Under Discussion
This KIP proposes an improvement that allows us to control whether we want to use instead of throwing away the prefetched data, simply return it along with the rest of the records coming from partitions that are not in paused state. This shall allow the client of the API to store that data and process it when ready instead of causing it to be fetched once again.
We add a boolean configuration parameter -
return.prefetched.data.for.paused, which defaults to
false. This will cause the
KafkaConsumer to return data that has already been prefetched even if the topic partition is in paused state..
The change being proposed here is simply using the configuration parameter in the
Fetcher class in order to opt out of aways throwing away prefetched data for partitions that are paused.to not throw the prefetched data and keep it around until the partition is resumed and polled again.
Compatibility, Deprecation, and Migration Plan
No breaking changes.