Uploaded image for project: 'WildFly'
  1. WildFly
  2. WFLY-13956

TwoConnectorsEJBFailoverTestCase sometimes hangs when running with jboss-ejb-client 4.0.35

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Done
    • Icon: Major Major
    • 22.0.0.Alpha1, 22.0.0.Final
    • 21.0.0.Beta1
    • EJB
    • None
    • Undefined
    • ---
    • ---

      When upgrading jboss-ejb-client to version 4.0.35.Final, we noticed some WildFly CI jobs sometimes hang when running TwoConnectorsEJBFailoverTestCase, causing 100+ (e.g., 466) tests to fail:

      RemoteElytronSingleSignOnTestCase.testSessionTimeoutDestroysSSOorg.jboss.as.test.clustering.cluster.sso.remote
RemoteElytronSingleSignOnTestCase.testFormAuthSingleSignOnorg.jboss.as.test.clustering.cluster.sso.remote
RemoteElytronSingleSignOnTestCase.testNoAuthSingleSignOnorg.jboss.as.test.clustering.cluster.sso.remote
RemoteSingleSignOnTestCase.testSessionTimeoutDestroysSSOorg.jboss.as.test.clustering.cluster.sso.remote
RemoteSingleSignOnTestCase.testFormAuthSingleSignOnorg.jboss.as.test.clustering.cluster.sso.remote
RemoteSingleSignOnTestCase.testNoAuthSingleSignOnorg.jboss.as.test.clustering.cluster.sso.remote
CoarseHotRodPersistenceWebFailoverTestCase.testGracefulUndeployFailoverorg.jboss.as.test.clustering.cluster.web.remote
CoarseHotRodPersistenceWebFailoverTestCase.testNonPrimaryOwnerorg.jboss.as.test.clustering.cluster.web.remote
CoarseHotRodPersistenceWebFailoverTestCase.testGracefulSimpleFailoverorg.jboss.as.test.clustering.cluster.web.remote
CoarseHotRodSessionActivationTestCase.testorg.jboss.as.test.clustering.cluster.web.remote
CoarseHotRodWebFailoverTestCase.testGracefulUndeployFailoverorg.jboss.as.test.clustering.cluster.web.remote
CoarseHotRodWebFailoverTestCase.testNonPrimaryOwnerorg.jboss.as.test.clustering.cluster.web.remote
CoarseHotRodWebFailoverTestCase.testGracefulSimpleFailoverorg.jboss.as.test.clustering.cluster.web.remote
CoarseTransactionalHotRodWebFailoverTestCase.testGracefulUndeployFailoverorg.jboss.as.test.clustering.cluster.web.remote
CoarseTransactionalHotRodWebFailoverTestCase.testNonPrimaryOwnerorg.jboss.as.test.clustering.cluster.web.remote
CoarseTransactionalHotRodWebFailoverTestCase.testGracefulSimpleFailoverorg.jboss.as.test.clustering.cluster.web.remote
FineHotRodPersistenceWebFailoverTestCase.testGracefulUndeployFailoverorg.jboss.as.test.clustering.cluster.web.remote

      The error messages are like:

      org.jboss.arquillian.container.spi.client.container.LifecycleException: The port 9990 is already in use. It means that either the server might be already running or there is another process using port 9990.
Managed containers do not support connecting to running server instances due to the possible harmful effect of connecting to the wrong server.
Please stop server (or another process) before running, change to another type of container (e.g. remote) or use jboss.socket.binding.port-offset variable to change the default port.
To disable this check and allow Arquillian to connect to a running server, set allowConnectingToRunningServer to true in the container configuration

      In a successful CI job, TwoConnectorsEJBFailoverTestCase runs 2 tests: testEJBClientUsingLegacyRemotingProtocol and testEJBClientUsingHttpUpgradeProtocol, each taking ~25 seconds to complete. See attached screenshot.

      But in a failed CI job, filtering by TwoConnectorsEJBFailoverTestCase gives no result, which means the test hangs at testEJBClientUsingLegacyRemotingProtocol and so the test runner hasn't been able to collect the result from either test.

        1. Screen Shot 2020-10-12 at 9.57.36 AM.png
          Screen Shot 2020-10-12 at 9.57.36 AM.png
          202 kB

              cfang@redhat.com Cheng Fang
              cfang@redhat.com Cheng Fang
              Votes:
              0 Vote for this issue
              Watchers:
              4 Start watching this issue

                Created:
                Updated:
                Resolved: