[
https://issues.apache.org/jira/browse/KAFKA-2589?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ewen Cheslack-Postava updated KAFKA-2589:
-----------------------------------------
Resolution: Fixed
Fix Version/s: 0.9.1.0
0.9.0.1
Status: Resolved (was: Patch Available)
Issue resolved by pull request 876
[https://github.com/apache/kafka/pull/876]
> Documentation bug: the default value for the "rebalance.backoff.ms" property
> is not specified correctly
> -------------------------------------------------------------------------------------------------------
>
> Key: KAFKA-2589
> URL: https://issues.apache.org/jira/browse/KAFKA-2589
> Project: Kafka
> Issue Type: Bug
> Components: config
> Affects Versions: 0.8.2.1
> Environment: any
> Reporter: Bogdan Dimitriu
> Assignee: Grant Henke
> Fix For: 0.9.0.1, 0.9.1.0
>
>
> The documentation for 0.8.2.1 consumer properties specifies:
> | rebalance.backoff.ms | 2000 | Backoff time between retries during rebalance
> |
> According to the source code though the default value is obtained this way:
> {code}
> val rebalanceBackoffMs = props.getInt("rebalance.backoff.ms", zkSyncTimeMs)
> {code}
> which is referenced from here:
> {code}
> val zkSyncTimeMs = props.getInt("zookeeper.sync.time.ms", 2000)
> {code}
> So by default it is 2000 as specified in the documentation, UNLESS the
> {{zookeeper.sync.time.ms}} is manually set to be different
> This may create confusion with recommendations such:
> {quote}
> In the case, make sure that rebalance.max.retries * rebalance.backoff.ms >
> zookeeper.session.timeout.ms
> {quote}
> from here:
> https://cwiki.apache.org/confluence/display/KAFKA/FAQ#FAQ-Whyaretheremanyrebalancesinmyconsumerlog?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
