Strange clustering behaviour

[prese]

Hi *,

We are using the following configuratioon:

- infinispan 4.1.0 GA

- jboss 6.0 M5(yes, we are nuts, but is the only jboss version working with almost the same jgroups version as infinispan)

We have configured inifinispan to be clusterd and to be part of the JTA transaction.

Duing the commit phase we are getting strange warnings and exceptions:

- Execution error: : org.infinispan.remoting.transport.jgroups.SuspectException: Suspected member: SebiPresecan-LT-26286

BUT my node is the ONLY node from the clusters. Thus it seems that he suspects himself. This is NOT REALLY FUNNY.

Another strange warning is: [org.jgroups.protocols.pbcast.GMS] (ViewHandler,SebiPresecan-LT-2069) I am the coord and I'm being suspected -- will probably leave shortly

We have used both the default transport protocol and the udp configuration. For both of them we have the same behaviour.

Any help is highly appreciated.

Thanks

[mircea.markus]

that's a bit strange indeed. Can you please enable TRACE for org.jgroups.protocols.pbcast.GMS and org.infinispan: this would hel us narrow down the problem. Also the jgroups and infinispan configuretion.

[prese]

Hi Mircea,

Please find attached the config files and the logs.

10x

Sebi

[mircea.markus]

The log shows that multiple caches are started. The good news is that you're not interfering with a coleague's cluster, all of them are started on your own machine :

[mmarkus:~/Downloads ]$ cat infinispan.log | grep "Cache local address is"

2010-09-30 16:22:36,811 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (2010-09-30T16:22:31,129 admin com.qnamic.railopt.project.command.LoadAreaCommand) Cache local address is SebiPresecan-LT-40678, physical addresses are [127.0.0.1:64566]

2010-09-30 16:22:39,506 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (2010-09-30T16:22:39,065 admin com.qnamic.railopt.plan.base.command.CreateChildPlanScheduleCommand) Cache local address is SebiPresecan-LT-21896, physical addresses are [127.0.0.1:64567]

2010-09-30 16:22:40,006 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (2010-09-30T16:22:39,629 admin com.qnamic.railopt.plan.base.command.CreateChildPlanScheduleCommand) Cache local address is SebiPresecan-LT-7515, physical addresses are [127.0.0.1:64568]

[mmarkus:~/Downloads ]$ cat infinispan.log | grep "Cache local address is"

2010-09-30 16:22:36,811 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (2010-09-30T16:22:31,129 admin com.qnamic.railopt.project.command.LoadAreaCommand) Cache local address is SebiPresecan-LT-40678, physical addresses are [127.0.0.1:64566]

2010-09-30 16:22:39,506 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (2010-09-30T16:22:39,065 admin com.qnamic.railopt.plan.base.command.CreateChildPlanScheduleCommand) Cache local address is SebiPresecan-LT-21896, physical addresses are [127.0.0.1:64567]

2010-09-30 16:22:40,006 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (2010-09-30T16:22:39,629 admin com.qnamic.railopt.plan.base.command.CreateChildPlanScheduleCommand) Cache local address is SebiPresecan-LT-7515, physical addresses are [127.0.0.1:64568]

Running UT at the same time perhaps?