5 Replies Latest reply on Jun 19, 2012 8:49 PM by roboticon

    fmc-webui bundle failure

    roboticon

      System Info

      - Using fmc 1.0.0-fuse-00-61 on Linux 64-bit with JDK 1.6...

      - fmc was installed from 32-bit Linux binary, no issues

      - A user and group was created to run fmc

       

      Error Info

      At FMC Console

      Running bin/fmc as the user created during the installation, I get an error on the console before the FMC graphic appears:

       

      Exception in thread "Thread-22" java.lang.ExceptionInInitializerError: org.linkedin.zookeeper.client.IZKClient service reference is null

         at com.fusesource.fmc.webui.Service$get_service(Services.scala:42)

         ...

       

      Log Exception

      using the command  "log:display-exception", I see

       

      java.lang.NullPointerException

        at com.sun.jersey.spi.container.ContainerRequest.(ContainerRequest.java:187)[135:com.fusesource.fmc.fmc_webui:1.0.0.-fuse-00-061]

       

      OSGi List

      Bundles with start id's 50, 53, 86, and 127 show Blueprint Failure status.  The start levels are 40, 45, 60, and 60 respectively

       

      Observables

      The FF browser shows a disabled "Client" button when browsing to http://localhost:8107/

       

      Caveats

      - The linux server does not have access to any maven repositories

       

      Thanks in advance for any help with this issue

       

      Edited by: roboticon on Jun 19, 2012 12:22 AM

        • 1. Re: fmc-webui bundle failure
          stlewis

          So the problem here is that fabric fails to start up one of it's services that's responsible for managing a fabric, the availability of that service controls whether or not FMC enables the "Create" button, as without that service you couldn't create a fabric anyway.

           

          Can you attach the log file from data/log/karaf.log?

           

          Also note you'll probably want to configure etc/org.ops4j.pax.url.mvn.cfg and add/set any internal maven proxy or server that you might be running, also worth probably creating etc/org.fusesource.fabric.maven.cfg and setting "remoteRepositories" to point to the same internal proxy, that way our maven proxy service can serve out artifacts to containers as you assign profiles on them.  The default value of this is:

           

          remoteRepositories=http://repo1.maven.org/maven2,http://repo.fusesource.com/nexus/content/groups/public,http://repo.fusesource.com/nexus/content/groups/ea

          • 2. Re: fmc-webui bundle failure
            stlewis

            Also, might be worth trying an osgi:refresh on any failed bundles to see if the bundles start up properly after the fact.

            • 3. Re: fmc-webui bundle failure
              roboticon

              Stan,

               

              I should probably have mentioned earlier that my system is on a private cloud (virtualized Linux servers) that is totally disconnected from the Internet (e.g. it does not have proxy access).

               

              Thanks,

              --scott c.

              • 4. Re: fmc-webui bundle failure
                roboticon

                Well, this is embarrassing...

                 

                While *typing in the contents of the data/karaf.log (as the machine has no internet connectivity) I noticed a

                 

                java.net.UnknownHostException

                 

                 

                This was the root of my problem.  All subsequent exceptions were eliminated by fixing my hostname table.

                 

                Thanks,

                --scott c.

                • 5. Re: fmc-webui bundle failure
                  roboticon

                  Well, this is embarrassing...

                   

                  While *typing in the contents of the data/karaf.log (as the machine has no internet connectivity) I noticed a

                   

                  java.net.UnknownHostException

                   

                   

                  This was the root of my problem.  All subsequent exceptions were eliminated by fixing my hostname table.

                   

                  Thanks,

                  --scott c.