Hi all, hope I'm in the right place for this.

I have a Tomcat 5.0 instance serving a couple of apps with fairly constant usage through the day. The hits are creeping upwards over time and about once a week for the last few weeks Tomcat has just stopped working. The logs show the applications working away and the next entry will be the restart.

Does anyone have any suggestions for monitoring what is happening? Tomcat is routinly restarted on the server each night to allow a refresh of some of the caching for the application, long standing issue that has no bearing on the issue apart from to say that Tomcat has only ever been running a maximum of a day when it happens.

While I'm at it can I recycle catalina.out as that is getting a bit big?

Any help on getting more clue to where the problem lies gratefully recieved.



By the way, this all runs on Fedora core 3.

Has the application closed down neatly through the destructor, or simply diasppeared?   If it's the former, you can perhaps add some trace code; it it's the latter, it's rather more awkward but I would be tempted to put in some sort of logging and look for patterns just before the failure.

Shouldn't be any problem recycling catalina.out; how about a crontab job that "mv"s the file to a separate logging directory, giving it a new name based on the day iof the week so that you have 7 old files to look back at?

Everything appears to just disappear.

Adding logging to the application is a problem due to the external testing and certification required. I am working on the assumption that either the system is running out of resources for some reason or the application is bringing down Tomcat rather than a problem with Tomcat itself. I guess I need to get improved logging into the application before the next testing phase.

Thanks for the help.

---