-
Bug
-
Resolution: Fixed
-
Minor
-
None
-
Jenkins 2.141 -(remoting 3.26)
kafka-remoting-plugin compiled against remoting 3.26
-
-
Remoting 3.28
I thought this was fixed in remoting 3.26 under https://issues.jenkins-ci.org/browse/JENKINS-42533
But today I got this stack trace again when one of my kafka nodes was idle (not running any job) just out of the blue it got
INFO: ConnectedINFO: Connected[Channel reader thread: planet] INFO org.apache.kafka.clients.Metadata - Cluster ID: isxhuhJhRGGNrKyNAeNZHg[Channel reader thread: planet] INFO org.apache.kafka.clients.consumer.internals.AbstractCoordinator - [Consumer clientId=consumer-2, groupId=jenkinssecure.8081.planet-id] Discovered group coordinator sydjenkikafka01.al.com.au:9092 (id: 2147483647 rack: null)[kafka-producer-network-thread | producer-2] INFO org.apache.kafka.clients.Metadata - Cluster ID: isxhuhJhRGGNrKyNAeNZHgOct 11, 2018 3:42:18 PM org.jenkinsci.remoting.util.AnonymousClassWarnings warnWARNING: Attempt to (de-)serialize anonymous class net.bull.javamelody.RemoteCallHelper$1; see: https://jenkins.io/redirect/serialization-of-anonymous-classes/ Oct 11, 2018 3:43:43 PM hudson.remoting.ExportTable unexportByOidS EVERE: Trying to unexport an object that's already unexportedjava.util.concurrent.ExecutionException: Invalid object ID 2 iota=3 at hudson.remoting.ExportTable.diagnoseInvalidObjectId(ExportTable.java:478) at hudson.remoting.ExportTable.unexportByOid(ExportTable.java:516) at hudson.remoting.Channel.unexport(Channel.java:803) at hudson.remoting.Channel.unexport(Channel.java:793) at hudson.remoting.UnexportCommand.execute(UnexportCommand.java:49) at hudson.remoting.Channel$1.handle(Channel.java:565) at hudson.remoting.SynchronousCommandTransport$ReaderThread.run(SynchronousCommandTransport.java:87)Caused by: java.lang.Exception: Object was recently deallocated #2 (ref.0) : object=null type=hudson.remoting.RemoteClassLoader$ClassLoaderProxy interfaces=[hudson.remoting.RemoteClassLoader$IClassLoader] Created at Thu Oct 11 15:42:13 AEDT 2018 at hudson.remoting.ExportTable$Entry.<init>(ExportTable.java:114) at hudson.remoting.ExportTable.export(ExportTable.java:365)
complete log attached.
Note the node did not stop working, at least I was able to run jobs after I noticed the stack trace.