[ https://issues.apache.org/jira/browse/KAFKA-4084?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17030183#comment-17030183 ]
GEORGE LI commented on KAFKA-4084: ---------------------------------- [~evanjpw] One solution we have in-house is a tooling to do "offline rebuild" of the failed empty broker. The idea is to keep the kafka on the new empty broker down. not serving any traffic. and copy the log segments/indexes files from other brokers in the cluster that this broker is supposed to be hosting. distribute and throttle the copy so it would not affect the brokers serving traffic in production. Once the copy is done, modify the replication offset meta file, start up the kafka on the new node, since there are historical logs already copied, the "delta" catch-up from leaders will be small, and it will be faster to catch up. This is just the main idea. It is an internal tooling that is tied to our infrastructure. Hope this helps. > 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)