You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@kafka.apache.org by "Kyle Ambroff-Kao (JIRA)" <ji...@apache.org> on 2018/01/23 03:26:00 UTC

[jira] [Created] (KAFKA-6468) Replication high watermark checkpoint file read for every LeaderAndIsrRequest

Kyle Ambroff-Kao created KAFKA-6468:
---------------------------------------

             Summary: Replication high watermark checkpoint file read for every LeaderAndIsrRequest
                 Key: KAFKA-6468
                 URL: https://issues.apache.org/jira/browse/KAFKA-6468
             Project: Kafka
          Issue Type: Bug
            Reporter: Kyle Ambroff-Kao


The high watermark for each partition in a given log directory is written to disk every _replica.high.watermark.checkpoint.interval.ms_ milliseconds. This checkpoint file is used to create replicas when joining the cluster.

[https://github.com/apache/kafka/blob/b73c765d7e172de4742a3aa023d5a0a4b7387247/core/src/main/scala/kafka/cluster/Partition.scala#L180]

Unfortunately this file is read every time kafka.cluster.Partition#getOrCreateReplica is invoked. For most clusters this isn't a big deal, but for a small cluster with lots of partitions all of the reads of this file really add up.

On my local test cluster of three brokers with around 40k partitions, the initial LeaderAndIsrRequest refers to every partition in the cluster, and it can take 20 to 30 minutes to create all of the replicas because the _replication-offset-checkpoint_ is nearly 2MB.

Changing this code so that we only read this file once on startup reduces the time to create all replicas to around one minute.

Credit to [~onurkaraman] for finding this one.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)