[
https://issues.apache.org/jira/browse/KAFKA-4084?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17086772#comment-17086772
]
GEORGE LI commented on KAFKA-4084:
----------------------------------
[~blodsbror]
I am not very familiar with 5.4 setup.
Do you have the error message of the crash in the log? is it missing the
zkclient jar like below?
{code}
$ ls -l zk*.jar
-rw-r--r-- 1 georgeli engineering 74589 Nov 18 18:21 zkclient-0.11.jar
$ jar tvf zkclient-0.11.jar
0 Mon Nov 18 18:11:58 UTC 2019 META-INF/
1135 Mon Nov 18 18:11:58 UTC 2019 META-INF/MANIFEST.MF
0 Mon Nov 18 18:11:58 UTC 2019 org/
0 Mon Nov 18 18:11:58 UTC 2019 org/I0Itec/
0 Mon Nov 18 18:11:58 UTC 2019 org/I0Itec/zkclient/
3486 Mon Nov 18 18:11:58 UTC 2019 org/I0Itec/zkclient/ContentWatcher.class
263 Mon Nov 18 18:11:58 UTC 2019 org/I0Itec/zkclient/DataUpdater.class
{code}
If this jar file was there before, please copy it back. I need to find out
why it was missing after the build. maybe some dependency setup in gradle. I
have also update the [install doc
|https://docs.google.com/document/d/14vlPkbaog_5Xdd-HB4vMRaQQ7Fq4SlxddULsvc3PlbY/edit]
using `./gradew clean build -x test`
Also make sure the startup script for kafka is not hard coding 5.4 jars, but
take the jars from the lib classpath? e.g.
{code}
/usr/lib/jvm/java-8-openjdk-amd64/bin/java
-Dlog4j.configuration=file:/etc/kafka/log4j.xml -Xms22G -Xmx22G -XX:+UseG1GC
-XX:MaxGCPauseMillis=20 -XX:NewSize=16G -XX:MaxNewSize=16G
-XX:InitiatingHeapOccupancyPercent=3 -XX:G1MixedGCCountTarget=1
-XX:G1HeapWastePercent=1 -XX:+PrintGCDetails -XX:+PrintGCTimeStamps
-XX:+PrintGCDateStamps -verbose:gc -Xloggc:/var/log/kafka/gc-kafka.log -server
-Dcom.sun.management.jmxremote
-Dcom.sun.management.jmxremote.authenticate=false
-Dcom.sun.management.jmxremote.ssl=false
-Dcom.sun.management.jmxremote.port=29010
-Djava.rmi.server.hostname=kafka12345-dca4 -cp '.:/usr/share/kafka/lib/*'
kafka.Kafka /etc/kafka/server.properties
{code}
If you give us more details, we can help more.
Thanks
> automated leader rebalance causes replication downtime for clusters with too
> many partitions
> --------------------------------------------------------------------------------------------
>
> Key: KAFKA-4084
> URL: https://issues.apache.org/jira/browse/KAFKA-4084
> Project: Kafka
> Issue Type: Bug
> Components: controller
> Affects Versions: 0.8.2.2, 0.9.0.0, 0.9.0.1, 0.10.0.0, 0.10.0.1
> Reporter: Tom Crayford
> Priority: Major
> Labels: reliability
> Fix For: 1.1.0
>
>
> If you enable {{auto.leader.rebalance.enable}} (which is on by default), and
> you have a cluster with many partitions, there is a severe amount of
> replication downtime following a restart. This causes
> `UnderReplicatedPartitions` to fire, and replication is paused.
> This is because the current automated leader rebalance mechanism changes
> leaders for *all* imbalanced partitions at once, instead of doing it
> gradually. This effectively stops all replica fetchers in the cluster
> (assuming there are enough imbalanced partitions), and restarts them. This
> can take minutes on busy clusters, during which no replication is happening
> and user data is at risk. Clients with {{acks=-1}} also see issues at this
> time, because replication is effectively stalled.
> To quote Todd Palino from the mailing list:
> bq. There is an admin CLI command to trigger the preferred replica election
> manually. There is also a broker configuration “auto.leader.rebalance.enable”
> which you can set to have the broker automatically perform the PLE when
> needed. DO NOT USE THIS OPTION. There are serious performance issues when
> doing so, especially on larger clusters. It needs some development work that
> has not been fully identified yet.
> This setting is extremely useful for smaller clusters, but with high
> partition counts causes the huge issues stated above.
> One potential fix could be adding a new configuration for the number of
> partitions to do automated leader rebalancing for at once, and *stop* once
> that number of leader rebalances are in flight, until they're done. There may
> be better mechanisms, and I'd love to hear if anybody has any ideas.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
