Re: [Wrapper-user] System.out deadlock

[Leif M. <le...@ta...>]

Andreas,
    Actually that is correct.   The problem was that up through version 
3.0.5, if a user
overrode the System.out or System.err print streams with their own and 
that code had
problems which lead to a dead lock or some other problem then the 
Wrapper was
being restarted in a way that made it look like a Wrapper bug.

    Another issue was that if the user redirected the streams then the 
Wrapper debug
output would be sent off to that other location as well as the user 
output.   If you are
trying to debug the Wrapper itself, you really want to keep all of the 
wrapper output
going to the wrapper.log file so that it can all be analyzed together.

    I modified the WrapperManager in version 3.1.0 so that it now 
remembers the
original System.out and System.err print streams and always uses that to 
output
its own output.

    If the user code hangs up, the Wrapper will still be able to 
successfully log its
own debug output, thus making it possible to track down the cause of the
problem.

    The way that test works is to enable debug output, then start the 
application.
Once the test app is up and running, I press the new "Poll Users" 
button. You will
start seeing user information interspersed with the debug output.  When 
you then
press "System.out Deadlock" button, the user information should stop because
it is using the user System.out object.  But the Wrapper's debug output 
should
still continue without any problems.  When you try to stop the 
application, this
will cause some output to go to the user System.out which is 
deadlocked.  This
will result in the application never shutting down cleanly so the 
Wrapper will
have to kill it.

    I'll have to add something about this to the docs...

Cheers,
Leif

Andreas Wendt wrote:

> Leif,
>
> when running the testwrapper, clicking on the "System.out deadlock" 
> button causes the JVM to hang without the wrapper recognizing it (at 
> least on MacOSX, OSF1, and Irix).
> What should really happen?
> And what is the purpose of this test? It is not explained in the manual.
>
> Cheers,
> Andreas
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> _______________________________________________
> Wrapper-user mailing list
> Wra...@li...
> https://lists.sourceforge.net/lists/listinfo/wrapper-user
>