Even more analysis (with dt_socket debug)
on one of our lab systems shows that it did loose
approx. 1000 threads overnite.

All of this Threads seem to have a target (runnable) of

org.hsqldb.Embeded_Server

These threads seem to indirectly be created by the 
LocalManagedConnectionFactory
when it obtains new connections,

(On this lab machine I use Hypersonic instead of Oracle. But the
Hypersonic Server is started on both, even if it is unused on
our production server with a similar leak...)

Half of the terminate, half of them persist (not running, not garbage 
collected).
I remember of a java bug like "ShutdownHook doubles thread count".
In my next tests I will disable the shutdownhook to see if there is some 
coincidence.

Regards,
Michael Bartmann





-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development

Reply via email to