Can't you catch Throwable as well and give us a stacktrace of whatever
exception/error gets thrown by the start() method?

regards,

Karl

On Wed, Sep 16, 2009 at 3:48 PM, Markus Michel
<[email protected]> wrote:
> Hi there!
>
> Here's a short explanation, how my program is working:
>
> Inside my program I using a FrameworkHandler class to control a pool of OSGi
> instances, which represents vehicles of a traffic simulation, For every
> instances I using three BundleHandler ( system, user, application) to
> install / uninstall / start / stop a list of bundles. When the program is
> started the pool is initialized and all system (OSGi services, Shell ..) and
> user (my services) bundles are installed and started. Everytime (if there is
> a free instance) there is a new vehicle, my program binds it to an instance
> of the pool and installs / started an application, which should be running
> on this vehicle using the application bundle handler. If the vehicles leaves
> the scenario, the program resets the application by stopping, uninstalled
> and reinstalling the corresponding bundles. Afterwards a new vehicle can be
> bound the same instance.
>
> My problem is that currently my program seems to hang into a loop, when the
> application bundle is started after it was reset. By doing some printouts I
> know that the problem occurs into that code block:
>
>            // start bundle
>            try
>            {
>                bundle.start();
>            }
>            catch (BundleException e)
>            {
>                System.out.println("error: could not start bundle " +
> bundle.getBundleId());
>                System.out.println(e);
>                System.exit(1);
>            }
>
> I'm not sure what really happens, because there is no bundle exception,
> although the error occurs inside the call of the start method. Isn't it
> possible to "reset" an OSGi bundle?
>
> Because I though that my problem might be a bug of the framework i already
> upgraded my program to the new felix version 2.0.0, but the problem still
> appears ...
>
> BR,
>
> Markus
>



-- 
Karl Pauls
[email protected]

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to