Details
-
Task
-
Status: Resolved
-
Major
-
Resolution: Won't Do
-
None
-
None
-
None
Description
When tiered storage is enabled on the cluster, Kafka broker has to build the remote log metadata for all the partitions that it is either leader/follower on node restart. The remote log metadata is built in asynchronous fashion and does not interfere with the broker startup path. Once the broker becomes online, it cannot handle the client requests (FETCH and LIST_OFFSETS) to access remote storage until the metadata gets built for those partitions. Currently, we are returning a ReplicaNotAvailable exception back to the client so that it will retry after sometime.
ReplicaNotAvailableException is applicable when there is a reassignment is in-progress and kind of deprecated with the NotLeaderOrFollowerException (PR#8979). It's good to introduce an appropriate retriable exception for remote storage errors to denote that it is not ready to accept the client requests yet.