[
https://issues.apache.org/jira/browse/KAFKA-2143?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14659503#comment-14659503
]
James Cheng commented on KAFKA-2143:
------------------------------------
Thanks [~becket_qin]. [~ewencp] sent me that slidedeck of yours. I added all
your suggested settings to my team's "Kafka recommended settings" wiki. :)
Even with ack=-1, I think it's still possible for a replica to lag behind the
leader by 2000 messages, right? It would be lagging behind by 2000 messages
that were produced but not-yet-acknowledged. Not sure if losing those would
count as message-loss, since they weren't actually ack'd?
replica.lag.max.messages is a broker-level config setting, not a topic level
config setting. I think that means that it's hard to find a setting that works
for high-volume topics and low-volume topics, right?
And... a bit of Googling around, and I think my observation is the exact thing
that inspired
https://cwiki.apache.org/confluence/display/KAFKA/KIP-16+-+Automated+Replica+Lag+Tuning
and
http://www.confluent.io/blog/hands-free-kafka-replication-a-lesson-in-operational-simplicity/,
right?
Is it possible for broker to be in the ISR for one topic, and to be out of the
ISR for another topic?
> Replicas get ahead of leader and fail
> -------------------------------------
>
> Key: KAFKA-2143
> URL: https://issues.apache.org/jira/browse/KAFKA-2143
> Project: Kafka
> Issue Type: Bug
> Components: replication
> Affects Versions: 0.8.2.1
> Reporter: Evan Huus
> Assignee: Jiangjie Qin
> Fix For: 0.8.3
>
>
> On a cluster of 6 nodes, we recently saw a case where a single
> under-replicated partition suddenly appeared, replication lag spiked, and
> network IO spiked. The cluster appeared to recover eventually on its own,
> Looking at the logs, the thing which failed was partition 7 of the topic
> {{background_queue}}. It had an ISR of 1,4,3 and its leader at the time was
> 3. Here are the interesting log lines:
> On node 3 (the leader):
> {noformat}
> [2015-04-23 16:50:05,879] ERROR [Replica Manager on Broker 3]: Error when
> processing fetch request for partition [background_queue,7] offset 3722949957
> from follower with correlation id 148185816. Possible cause: Request for
> offset 3722949957 but we only have log segments in the range 3648049863 to
> 3722949955. (kafka.server.ReplicaManager)
> [2015-04-23 16:50:05,879] ERROR [Replica Manager on Broker 3]: Error when
> processing fetch request for partition [background_queue,7] offset 3722949957
> from follower with correlation id 156007054. Possible cause: Request for
> offset 3722949957 but we only have log segments in the range 3648049863 to
> 3722949955. (kafka.server.ReplicaManager)
> [2015-04-23 16:50:13,960] INFO Partition [background_queue,7] on broker 3:
> Shrinking ISR for partition [background_queue,7] from 1,4,3 to 3
> (kafka.cluster.Partition)
> {noformat}
> Note that both replicas suddenly asked for an offset *ahead* of the available
> offsets.
> And on nodes 1 and 4 (the replicas) many occurrences of the following:
> {noformat}
> [2015-04-23 16:50:05,935] INFO Scheduling log segment 3648049863 for log
> background_queue-7 for deletion. (kafka.log.Log) (edited)
> {noformat}
> Based on my reading, this looks like the replicas somehow got *ahead* of the
> leader, asked for an invalid offset, got confused, and re-replicated the
> entire topic from scratch to recover (this matches our network graphs, which
> show 3 sending a bunch of data to 1 and 4).
> Taking a stab in the dark at the cause, there appears to be a race condition
> where replicas can receive a new offset before the leader has committed it
> and is ready to replicate?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
