Required Data: HA (active/active) Hazelcast EventQueue overloaded

1 minute readKnowledge base

Issue

In your controller logs of an HA (active/active) cluster, you see the following error message:

EventQueue overloaded! Failed to process event packet sent from:

Resolution

  1. To troubleshoot this when running CloudBees CI version 2.452.3.2 or newer:

    1. Collect a support bundle with the About Jenkins, Information from other replicas, and Hazelcast collection summary options.

    2. The support bundle will contain information about the Hazelcast collections in the file hazelcast-collections.txt.

  2. To troubleshoot this when running CloudBees CI versions older than 2.452.3.2:

    1. Add a FINER level logger for com.cloudbees.jenkins.plugins.replication.sync by following Configure Loggers for Jenkins and CloudBees CI. This will provide more information about the events being processed.

      FINER c.c.j.p.r.s.MessageProvider$ListenerImpl#isActuallyModified: checksum of ... changed so considering save event
    2. Run collectPerformanceData.sh to get consecutive thread dumps when the error is happening, by following Required Data: CloudBees CI hang or high CPU on Kubernetes or Linux.

    3. Collect a support bundle with the About Jenkins, Information from other replicas, and Controller Custom Log Recorders options.

  3. Submit the above data to CloudBees Support for further analysis at https://support.cloudbees.com.

This article is part of our Knowledge Base and is provided for guidance-based purposes only. The solutions or workarounds described here are not officially supported by CloudBees and may not be applicable in all environments. Use at your own discretion, and test changes in a safe environment before applying them to production systems.