Hide Forgot
project_key: EDG infinispan version: 4.2.1.CR3 This appeared in the server log while running performance test (client stress test on 6 node cluster): java.lang.UnsupportedOperationException: Unsupported! [JBoss] at org.infinispan.distribution.ch.UnionConsistentHash.getHashId(UnionConsistentHash.java:53) [JBoss] at org.infinispan.server.hotrod.HotRodEncoder$$anonfun$writeHashTopologyHeader$4.apply(HotRodEncoder.scala:164) [JBoss] at org.infinispan.server.hotrod.HotRodEncoder$$anonfun$writeHashTopologyHeader$4.apply(HotRodEncoder.scala:160) [JBoss] at scala.collection.LinearSeqOptimized$class.foreach(LinearSeqOptimized.scala:61) [JBoss] at scala.collection.immutable.List.foreach(List.scala:45) [JBoss] at org.infinispan.server.hotrod.HotRodEncoder.writeHashTopologyHeader(HotRodEncoder.scala:160) [JBoss] at org.infinispan.server.hotrod.HotRodEncoder.writeHeader(HotRodEncoder.scala:117) [JBoss] at org.infinispan.server.hotrod.HotRodEncoder.encode(HotRodEncoder.scala:32) [JBoss] at org.infinispan.server.core.transport.netty.EncoderAdapter.encode(EncoderAdapter.scala:17) [JBoss] at org.jboss.netty.handler.codec.oneone.OneToOneEncoder.handleDownstream(OneToOneEncoder.java:66) [JBoss] at org.jboss.netty.channel.Channels.write(Channels.java:611) [JBoss] at org.jboss.netty.channel.Channels.write(Channels.java:578) [JBoss] at org.jboss.netty.channel.AbstractChannel.write(AbstractChannel.java:258) [JBoss] at org.infinispan.server.core.transport.netty.ChannelAdapter.write(ChannelAdapter.scala:21) [JBoss] at org.infinispan.server.core.AbstractProtocolDecoder.writeResponse(AbstractProtocolDecoder.scala:87) [JBoss] at org.infinispan.server.core.AbstractProtocolDecoder.decode(AbstractProtocolDecoder.scala:51) [JBoss] at org.infinispan.server.core.transport.netty.DecoderAdapter.decode(DecoderAdapter.scala:18) [JBoss] at org.infinispan.server.core.transport.netty.DecoderAdapter.decode(DecoderAdapter.scala:14) [JBoss] at org.jboss.netty.handler.codec.replay.ReplayingDecoder.callDecode(ReplayingDecoder.java:461) [JBoss] at org.jboss.netty.handler.codec.replay.ReplayingDecoder.messageReceived(ReplayingDecoder.java:434) [JBoss] at org.jboss.netty.channel.Channels.fireMessageReceived(Channels.java:274) [JBoss] at org.jboss.netty.channel.Channels.fireMessageReceived(Channels.java:261) [JBoss] at org.jboss.netty.channel.socket.nio.NioWorker.read(NioWorker.java:349) [JBoss] at org.jboss.netty.channel.socket.nio.NioWorker.processSelectedKeys(NioWorker.java:281) [JBoss] at org.jboss.netty.channel.socket.nio.NioWorker.run(NioWorker.java:201) [JBoss] at org.jboss.netty.util.internal.IoWorkerRunnable.run(IoWorkerRunnable.java:46) [JBoss] at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) [JBoss] at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) [JBoss] at java.lang.Thread.run(Thread.java:619) full log: http://hudson.qa.jboss.com/hudson/view/EDG/job/edg-51x-stress-client-size6-hotrod/1/console-perf01/consoleText this will be probably hard to reproduce, but I guess is worth reporting and investigating.
Hmmm, this is something that needs looking into. That UnionConsistentHash is used when rehashing is going on and a new view is being installed. It appears that the hash id for a given node cannot be calculated at that point. I'll probably need to check with Manik on this. Good catch Michal!
Link: Added: This issue depends ISPN-963
today Rado also ran into this problem when running a job: http://hudson.qa.jboss.com/hudson/view/EDG/job/edg-51x-resilience-client-size4-hotrod/6/console-perf02/
Michal/Rado, this has now been fixed and we'll release today a new CR (March 7). Can you make sure you run your tests to verify it works fine now? Marking it as resolved for the time being.
Michal, CR4 is out now. Make sure you integrate and test it asap since we're targeting a Final at the end of this week.
Sure, we're on it.
Well, I've updated EDG to use infinispan 4.2.1.CR4 and Rado has rerun his job, we saw these exceptions in. They are not occuring, but I don't know whether this was a valid run. http://hudson.qa.jboss.com/hudson/view/EDG/job/edg-51x-resilience-client-size4-hotrod/13/ It's not currently possible to rerun my original job http://hudson.qa.jboss.com/hudson/view/EDG/job/edg-51x-stress-client-size6-hotrod/ it was only a short time for which I had perf lab machines available for size6 cluster tests. It might take a little longer till we can run such jobs again. And even then It occured only once in my run. Rado had multiple occurences of them. Since Rado is already gone, we'll try to rerun Rado's job tomorrow and verify again.
Rado did another runs of his job and we didn't see the exception again http://hudson.qa.jboss.com/hudson/view/EDG/job/edg-51x-resilience-client-size4-hotrod/ I don't know how to exactly replicate this so this so we didn't verify the issue for 100%, but it seems that it's ok now. I'll run one more test with CR3 to see whether we'll see the exceptions there.
unfortunately the same run with 4.2.1.CR3 didn't show the exceptions.
Michal/Rado, just keep running your tests and if it appears again, we'll look into it.
Link: Added: This issue is a dependency of JBPAPP-6054