You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@kafka.apache.org by "Luke Chen (Jira)" <ji...@apache.org> on 2025/03/06 11:04:00 UTC

[jira] [Comment Edited] (KAFKA-18871) KRaft migration rollback causes downtime

    [ https://issues.apache.org/jira/browse/KAFKA-18871?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17932941#comment-17932941 ] 

Luke Chen edited comment on KAFKA-18871 at 3/6/25 11:03 AM:
------------------------------------------------------------

[~davidarthur] , thanks for the comment.

> Even though the write back to ZK is asynchronous, we should still be able to ensure consistency by cleanly shutting down the controllers. The event queue should process any remaining events when doing a normal shutdown of the controller nodes. With all of the controllers shut down, the metadata won't be able to change any further and it is safe to elect a ZK controller.

 

Unfortunately, our current implementation is not able to ensure consistency between ZK/KRaft after cleanly shutdown. I've filed KAFKA-18930 for detailed description. I agree with you that we have to guarantee the consistency between ZK/KRaft when KRaft is cleanly shutdown.

> A workaround might be to break the network between the brokers and KRaft controllers to allow the metadata to sync before taking the controllers down.

 

I think if we don't retry the KRaft -> ZK write, they won't sync eventually.

 

Before KAFKA-18930  is fixed, I'd like to propose we add some notes in 3.9 KRaft migration section about this known issue. WDYT?


was (Author: showuon):
[~davidarthur] , thanks for the comment.

> Even though the write back to ZK is asynchronous, we should still be able to ensure consistency by cleanly shutting down the controllers. The event queue should process any remaining events when doing a normal shutdown of the controller nodes. With all of the controllers shut down, the metadata won't be able to change any further and it is safe to elect a ZK controller.

 

Unfortunately, our current implementation is not able to ensure consistency between ZK/KRaft after cleanly shutdown. I've filed KAFKA-18930 for detailed description. I agree with you that we have to guarantee the consistency between ZK/KRaft when KRaft is cleanly shutdown.

> KRaft migration rollback causes downtime
> ----------------------------------------
>
>                 Key: KAFKA-18871
>                 URL: https://issues.apache.org/jira/browse/KAFKA-18871
>             Project: Kafka
>          Issue Type: Bug
>          Components: kraft, migration
>    Affects Versions: 3.9.0
>            Reporter: Daniel Urban
>            Priority: Critical
>         Attachments: cluster-operator.log.zip, controller_logs.zip, kraft-rollback-bug.zip, kraft-rollback-kafka-default-pool-0.zip, kraft-rollback-kafka-default-pool-1.zip, kraft-rollback-kafka-default-pool-2.zip, strimzi-cluster-operator-7bc47d488f-p4ltv.zip
>
>
> When testing the KRaft migration rollback feature, found the following scenario:
>  # Execute KRaft migration on a 3 broker 3 ZK node cluster to the last step, but do not finalize the migration.
>  ## In the test, we put a slow but continuous produce+consume load on the cluster, with a topic (partitions=3, RF=3, min ISR=2)
>  # Start the rollback procedure
>  # First we roll back the brokers from KRaft mode to migration mode (both controller and ZK configs are set, process roles are removed, {{zookeeper.metadata.migration.enable}} is true)
>  # Then we delete the KRaft controllers, delete the /controller znode
>  # Then we immediately start rolling the brokers 1 by 1 to ZK mode by removing the {{zookeeper.metadata.migration.enable}} flag and the controller.* configurations.
>  # At this point, when we restart the 1st broker (let's call it broker 0) into ZK mode, we find an issue which occurs ~1 out of 20 times:
> If broker 0 is not in the ISR for one of the partitions at this point, it can simply never become part of the ISR. Since we are aiming for zero downtime, we check the ISR states of partitions between broker restarts, and our process gets blocked at this point. We have tried multiple workarounds at this point, but it seems that there is no workaround which still ensures zero downtime.
> Some more details about the process:
>  * We are using Strimzi to drive this process, but have verified that Strimzi follows the documented steps precisely.
>  * When we reach the error state, it doesn't matter which broker became the controller through the ZK node, the brokers still in migration mode get stuck, and they flood the logs with the following error:
> {code:java}
> 2025-02-26 10:55:21,985 WARN [RaftManager id=0] Error connecting to node kraft-rollback-kafka-controller-pool-5.kraft-rollback-kafka-kafka-brokers.csm-op-test-kraft-rollback-e7798bef.svc.cluster.local:9090 (id: 5 rack: null) (org.apache.kafka.clients.NetworkClient) [kafka-raft-outbound-request-thread]
> java.net.UnknownHostException: kraft-rollback-kafka-controller-pool-5.kraft-rollback-kafka-kafka-brokers.csm-op-test-kraft-rollback-e7798bef.svc.cluster.local
>         at java.base/java.net.InetAddress$CachedAddresses.get(InetAddress.java:801)
>         at java.base/java.net.InetAddress.getAllByName0(InetAddress.java:1533)
>         at java.base/java.net.InetAddress.getAllByName(InetAddress.java:1385)
>         at java.base/java.net.InetAddress.getAllByName(InetAddress.java:1306)
>         at org.apache.kafka.clients.DefaultHostResolver.resolve(DefaultHostResolver.java:27)
>         at org.apache.kafka.clients.ClientUtils.resolve(ClientUtils.java:125)
>         at org.apache.kafka.clients.ClusterConnectionStates$NodeConnectionState.resolveAddresses(ClusterConnectionStates.java:536)
>         at org.apache.kafka.clients.ClusterConnectionStates$NodeConnectionState.currentAddress(ClusterConnectionStates.java:511)
>         at org.apache.kafka.clients.ClusterConnectionStates$NodeConnectionState.access$200(ClusterConnectionStates.java:466)
>         at org.apache.kafka.clients.ClusterConnectionStates.currentAddress(ClusterConnectionStates.java:173)
>         at org.apache.kafka.clients.NetworkClient.initiateConnect(NetworkClient.java:1075)
>         at org.apache.kafka.clients.NetworkClient.ready(NetworkClient.java:321)
>         at org.apache.kafka.server.util.InterBrokerSendThread.sendRequests(InterBrokerSendThread.java:146)
>         at org.apache.kafka.server.util.InterBrokerSendThread.pollOnce(InterBrokerSendThread.java:109)
>         at org.apache.kafka.server.util.InterBrokerSendThread.doWork(InterBrokerSendThread.java:137)
>         at org.apache.kafka.server.util.ShutdownableThread.run(ShutdownableThread.java:136) {code}
>  * Manually verified the last offsets of the replicas, and broker 0 is caught up in the partition.
>  * Even after stopping the produce load, the issue persists.
>  * Even after removing the /controller node manually (to retrigger election), regardless of which broker becomes the controller, the issue persists.
> Based on the above, it seems that during the rollback, brokers in migration mode cannot handle the KRaft controllers being removed from the system. Since broker 0 is caught up in the partition, we suspect that the other brokers (still in migration mode) do not respect the controller state in ZK, and do not report changes in the ISR of the partitions they are leading.
> This means that if a replica becomes out of sync in the last restart (e.g. due to a slow broker restart), we cannot restart the brokers while ensuring zero downtime.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)