[
https://issues.apache.org/jira/browse/KAFKA-20788?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Aditya Kousik updated KAFKA-20788:
----------------------------------
Description:
A Kafka consumer client today cannot tell why a rebalance happened without
grepping logs. The rebalance trigger can help with triaging:
* {{max.poll.interval.ms}} exceeded implies misconfig or delayed poll loop;
* Member fencing points at session timeout
* Coordinator-initiated rebalances are regular occurrences due to join and
leave requests.
* Application calls to {{Consumer.enforceRebalance()}} are expected and the
application owns the context.
Existing rebalance metrics describe frequency and latency but carry no cause
dimension:
{\{rebalance-rate-per-hour, rebalance-latency-{avg,max}, rebalance-total
last-rebalance-seconds-ago, last-poll-seconds-ago}}
A spike in {{rebalance-rate-per-hour}} cannot be narrowed without consumer logs.
KIP:
[https://cwiki.apache.org/confluence/spaces/KAFKA/pages/440304013/KIP-1364+Expose+Rebalance+Cause+to+RebalanceConsumer+and+Metrics]
was:
A Kafka consumer client today cannot tell why a rebalance happened without
grepping logs. The rebalance trigger can help with triaging:
* {{max.poll.interval.ms}} exceeded implies misconfig or delayed poll loop;
* Member fencing points at session timeout
* Coordinator-initiated rebalances are regular occurrences due to join and
leave requests.
* Application calls to {{Consumer.enforceRebalance()}} are expected and the
application owns the context.
Existing rebalance metrics describe frequency and latency but carry no cause
dimension:
{{rebalance-rate-per-hour, rebalance-latency-\{avg,max}, rebalance-total
last-rebalance-seconds-ago, last-poll-seconds-ago}}
A spike in {{rebalance-rate-per-hour}} cannot be narrowed without consumer logs.
> Expose Rebalance Cause to RebalanceConsumer and Metrics
> -------------------------------------------------------
>
> Key: KAFKA-20788
> URL: https://issues.apache.org/jira/browse/KAFKA-20788
> Project: Kafka
> Issue Type: New Feature
> Components: clients
> Reporter: Aditya Kousik
> Assignee: Aditya Kousik
> Priority: Major
> Labels: kip
>
> A Kafka consumer client today cannot tell why a rebalance happened without
> grepping logs. The rebalance trigger can help with triaging:
> * {{max.poll.interval.ms}} exceeded implies misconfig or delayed poll loop;
> * Member fencing points at session timeout
> * Coordinator-initiated rebalances are regular occurrences due to join and
> leave requests.
> * Application calls to {{Consumer.enforceRebalance()}} are expected and the
> application owns the context.
> Existing rebalance metrics describe frequency and latency but carry no cause
> dimension:
> {\{rebalance-rate-per-hour, rebalance-latency-{avg,max}, rebalance-total
> last-rebalance-seconds-ago, last-poll-seconds-ago}}
> A spike in {{rebalance-rate-per-hour}} cannot be narrowed without consumer
> logs.
>
> KIP:
> [https://cwiki.apache.org/confluence/spaces/KAFKA/pages/440304013/KIP-1364+Expose+Rebalance+Cause+to+RebalanceConsumer+and+Metrics]
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
