IceGridAdmin exceptions

[ctennis]

I've been doing a lot of adding/removing and maintenance of IceGrid via IceGridAdmin. One exception I'm running into a lot gives this backtrace:

java.lang.NullPointerException
at IceGridGUI.LiveDeployment.Node.removeDescriptor(No de.java:890)
at IceGridGUI.LiveDeployment.Node.update(Node.java:38 4)
at IceGridGUI.LiveDeployment.Root.applicationUpdated( Root.java:481)
at IceGridGUI.Coordinator.applicationUpdated(Coordina tor.java:755)
at IceGridGUI.ApplicationObserverI$3.run(ApplicationO bserverI.java:140)
at java.awt.event.InvocationEvent.dispatch(Unknown Source)
at java.awt.EventQueue.dispatchEvent(Unknown Source)
at java.awt.EventDispatchThread.pumpOneEventForHierar chy(Unknown Source)
at java.awt.EventDispatchThread.pumpEventsForHierarch y(Unknown Source)
at java.awt.EventDispatchThread.pumpEvents(Unknown Source)
at java.awt.EventDispatchThread.pumpEvents(Unknown Source)
at java.awt.EventDispatchThread.run(Unknown Source)

I don't know if I can describe what I'm doing that's causing it, other than just saving the properties to the registry. I must be putting it into some state it doesn't want to be in before that point.

Anyway, thought this might be of some value.

[bernard]

Hi Caleb,

Thanks for the bug report; this stack trace shows that the registry tells IceGrid Admin to remove a server that it does not know about -- which should not happen. After you get this error, I'd recommend to reconnect to the registry (or restart IceGrid Admin), to ensure that IceGrid Admin's local data structures are correct.

Did you get any other exception stack trace before this one? The first stack trace is the most useful one.

Thanks,
Bernard

[ctennis]

This is the only stack trace I receive. When it happens, IceGridAdmin continues to run, but it has a stopwatch mousecursor and will only let me click on a few items - one of which is the disconnection button. Disconnecting and reconnecting shows I'm in a consistent state with what I expected.

This came about when changing some Server Template properties and Well Known Objects (specifically, changing template things like ${index} to ${node} and such), which affected some already deployed (but not running) servers.

I also had another instance of IceGridAdmin running on another machine, and these exceptions showed up in the stderr of that machine as well. Disconnecting and reconnecting seemed to work just fine.

Anyway, I'm past it all now, but I thought it might be interesting to note.