[KAFKA-10879] ReplicaFetcherThread crash when cluster doing reassign

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 2.1.1
• Component/s: replication
• Assignee: Unassigned
• Reporter: zhifeng.peng
• Created: 22/Dec/20 08:04
• Updated: 23/Dec/20 08:10

Description

[2020-12-21 12:01:28,110] ERROR [ReplicaFetcher replicaId=3, leaderId=6, fetcherId=9] Error due to (kafka.server.ReplicaFetcherThread)
java.lang.NullPointerException
[2020-12-21 12:01:28,110] INFO [ReplicaFetcher replicaId=3, leaderId=6, fetcherId=9] Stopped (kafka.server.ReplicaFetcherThread)

During reassign partition, one partition was stucked. I find one ReplicaFetcherThread has stoped. It resulting in some partition under-Replicated. I have attach jstack infomation.