Re: Just me? Debugged JVM's hanging, preventing IPL?

David Gibbs skrev den 22-05-2008 15:40:

Thorbjørn Ravn Andersen wrote:

I have had a problem for quite some time when debugging Java programs on V5R3M0 from Eclipse on PC, namely that once in a while the JVM gives a strange message and then hangs. It does not take any notice of me trying to end the job, either with SysReq-2 *I or with the 4 option in wrkactjob. It just sits there (the one I have now is in THDW).

Well, I've never actually debugged a java program on the i ... I always do my development & debugging on a PC (which, when I thought about it, surprised the heck out of me because I've never had a platform specific problem with the java code).

This particular program I have inherited and it is intended to run on the i, so it has many assumptions that I have not yet refactored out (i.e. THIS File object corresponds to THAT member etc). It is easier to just run it on the i as I need to finish this rather soonish. I deploy the jar, invoke the program, attach the debugger - it is ok as long as it does not crash :).
But do everybody else do it on their PC and just deploy in production on the i?

Anyways ... has the code ever had a similar problem when running off the i?

Code runs fine when not running in debug mode. Sigh.

This means that it fires up the JVM and wait for me to attach with the debugger before starting the program. It then runs and once in a while, in an inreproducible way to me, complains and hang.

What kind of complaint?

Varies a bit, but today the Eclipse debugger said that the backend returned something strange with an error code, and I didn't think of writing it down.

Obvious question ... are you current only java group ptf's?

I believe that we are fully up to date.