[
https://issues.apache.org/jira/browse/KAFKA-7040?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16508847#comment-16508847
]
Dong Lin commented on KAFKA-7040:
---------------------------------
[~luwang] Nice catch. It is indeed problematic (with some probability of losing
data) that a replica fetcher thread can truncate to an offset smaller than the
high watermark due to receiving an "outdated" OffsetsForLeaderEpochResponse.
One possible solution is to add one more field, say `initialized`, in
AbstractFetcherThread.partitionStates. The value is set to false after the
partition is added to partitionStates. ReplicaFetcherThread will update
`initialized` to true the first time it reads this partition. Later, after
ReplicaFetcherThread receives either FetchResponse or
OffsetsForLeaderEpochResponse, it should discard this partitions in the
response whose `initialized` is false. This seems to fix the issue here.
> The replica fetcher thread may truncate accepted messages during multiple
> fast leadership transitions
> -----------------------------------------------------------------------------------------------------
>
> Key: KAFKA-7040
> URL: https://issues.apache.org/jira/browse/KAFKA-7040
> Project: Kafka
> Issue Type: Bug
> Reporter: Lucas Wang
> Priority: Minor
>
> Problem Statement:
> Consider the scenario where there are two brokers, broker0, and broker1, and
> there are two partitions "t1p0", and "t1p1"[1], both of which have broker1 as
> the leader and broker0 as the follower. The following sequence of events
> happened on broker0
> 1. The replica fetcher thread on a broker0 issues a LeaderEpoch request to
> broker1, and awaits to get the response
> 2. A LeaderAndISR request causes broker0 to become the leader for one
> partition t1p0, which in turn will remove the partition t1p0 from the replica
> fetcher thread
> 3. Broker0 accepts some messages from a producer
> 4. A 2nd LeaderAndISR request causes broker1 to become the leader, and
> broker0 to become the follower for partition t1p0. This will cause the
> partition t1p0 to be added back to the replica fetcher thread on broker0.
> 5. The replica fetcher thread on broker0 receives a response for the
> LeaderEpoch request issued in step 1, and truncates the accepted messages in
> step3.
> The issue can be reproduced with the test from
> https://github.com/gitlw/kafka/commit/8956e743f0e432cc05648da08c81fc1167b31bea
> [1] Initially we set up broker0 to be the follower of two partitions instead
> of just one, to avoid the shutting down of the replica fetcher thread when it
> becomes idle.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
