Re: [certi-dev] Deadlock with jcerti

[Eric Noulard]

2012/10/15 Jan-Patrick Osterloh <address@hidden>:
> Hi all,
>
> here is the patch. I added a class in the logging package for reading
> the streams and putting it into the logger. You may want to change the
> logger or the logging function. This class is then attached to the RTIA
> process, which is created in the CertiRtiaAmbassador.

The patch looks ok to me I'll test it soon.

Your analysis of the problem is sound.
I did manage to reproduce the lock with RTIA=D in the environment
but I had to wait for a couple of minute before deadlock.

This should heavily depends on default buffer size of TCP socket
(should be 16ko on linux for ipv4).

Thanks for investigating this and for the patch.

By the way did the patch works as expected for you?
i.e. no more lock?

-- 
Erk
Le gouvernement représentatif n'est pas la démocratie --
http://www.le-message.org