Ever since I added JCS caching to Hibernate 2.1 my Tomcat server likes to die 6-12 hours after it was started up. Absolutely no logging messages are written to catalina.out, the host logger, or a log4j log, the server just stops.

I figure this is probably not a problem with Hibernate but I was wondering if anyone had experience using JCS and might have experienced something like this. I'm aware that some people don't think JCS is a very good caching package though perhaps it is the best free one. If anyone knows how to tune JCS such that it can keep going for days at a time, please make some suggestions.

I'm using Sun's JDK 1.4.2_01 on RedHat 9. I'll try the _02 release that came out this week and see if that changes anything. I'm also interested in any suggestions about diagnosing the problem further. I don't really know that much about Linux, where crash logs might be located, etc. This probably is obviously hard to test/diagnose because it only happens twice a day at most.

Thanks,

I just committed the new oscache and swarmcache patches to CVS. Wanna be the guinea pig?

I'll try it out at the next release. We are actually in production so as much as I'd like to be your pig, it's probably not a responsible decision.

Issues covered in more detail here:

http://forum.hibernate.org/viewtopic.php?t=924937