AMQ broker pod blocked when split folder is locked by AMQ Drainer pod
Issue
- When we start a new deployment of AMQ broker pod, if this take long time, the drainer pod take the lock of
split-1
folder and the amq broker pod tried with the split folder locked by the AMQ drainer pod, then shows the following messages in console and keeps blocked at this stage:
-Xms1024m -Xmx1024m -XX:+UseParallelGC -XX:MinHeapFreeRatio=20 -XX:MaxHeapFreeRatio=40 -
XX:GCTimeRatio=4 -XX:AdaptiveSizePolicyWeight=90 -XX:MaxMetaspaceSize=100m -XX:ParallelGCThreads=1 -
Djava.util.concurrent.ForkJoinPool.common.parallelism=1 -XX:CICompilerCount=2 -XX:+ExitOnOutOfMemoryError
Using jmxtrans agent to collect metrics. Configuration loaded from /opt/amq/conf/jmxtrans-agent-kafka-influxdb.xml
Running jboss-amq-6/amq63-openshift image, version 1.3
Attempting to obtain lock for directory: (/opt/amq/data/split-1)
Attempting to obtain lock for directory: (/opt/amq/data/split-2)
Existing server instance is terminating, waiting to acquire the lock
Environment
- OpenShift : 3.9
- XPaaS AMQ : jboss-amq-6/amq63-openshift:1.3
- GlusterFS
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.