Well, in details it means that is trying to write/delete at same key and it failed, and then probably the indexmanager was not updated.

Did you get that trace also in the server.log as well ?

On Mon, Oct 30, 2017 at 6:25 PM, Lucas Ponce <lponce@redhat.com> wrote:
That's interesting.
I haven't got that trace, yes, it seems the indexmanager failed, still trying to get the same one.

On Mon, Oct 30, 2017 at 6:17 PM, Joel Takvorian <jtakvori@redhat.com> wrote:
Some update, here is the latest logs I got while the test failed:

I haven't time to dig further today, but see in the middle that lovely stack trace:

18:01:59,012 ERROR [org.infinispan.transaction.impl.TransactionCoordinator] (default task-35) ISPN000255: Error while processing prepare: org.infinispan.transaction.WriteSkewException: Write skew detected on key itest-rest-feed~Local DMR~/subsystem=datasources/xa-data-source=testXaDs for transaction TransactionImple < ac, BasicAction: 0:ffffc0a8000c:-5784e282:59f75acd:b4 status: ActionStatus.RUNNING >
at org.infinispan.transaction.impl.WriteSkewHelper.performWriteSkewCheckAndReturnNewVersions(WriteSkewHelper.java:66)
at org.infinispan.interceptors.locking.ClusteringDependentLogic$AbstractClusteringDependentLogic.clusteredCreateNewVersionsAndCheckForWriteSkews(ClusteringDependentLogic.java:261)

I have no idea why it wasn't displayed before, and it is now. 

On Mon, Oct 30, 2017 at 1:15 PM, Joel Takvorian <jtakvori@redhat.com> wrote:
I've seen the same error just once (out of 5 or 6 tries), and, coincidentally, had the same error notified "Failed to store notification: {"type":"RESOURCE_ADDED" etc." with "MethodNotAllowed" at the very same moment, that is, just when the testRemoveXaDatasource test runs.
That's maybe a trail to investigate, maybe it betrays some server-side malfunction. I don't know if it's possible to see server-side logs during agent itests, but it would probably help. I don't know why this error happens.

When I run tests successfully, I also see from times to times this error, but it's distributed randomly in tests, and fortunately or not (rather not), tests pass.


_______________________________________________
hawkular-dev mailing list
hawkular-dev@lists.jboss.org
https://lists.jboss.org/mailman/listinfo/hawkular-dev