since 10/21 we've had tomcat crash six times on an internet facing novell-groupwise-webaccess-8.0.0HP-87328 system running sles 10. apache is still up, and repeatedly records the following messages in its error_log:

[error] (104)Connection reset by peer: ajp_ilink_receive() can't receive header

customers' browsers start displaying a "Service unavailable!" page from apache: "The server is temporarily unable to service your request due to maintenance downtime or capacity problems. Please try again later. If you think this is a server error, please contact the webmaster. Error 503 Apache/2.2.3 (Linux/SUSE)". a tech must start tomcat5 to bring webaccess back online.

only change has been user load; pre-production had been hovering around ~800 hits/day, and after the 10/20 go-live, we've seen ~5000 hits/day. max concurrency has been ~100.

tomcat5/apache2 installed on this one system in dmz, with its gw agent domain on separate system behind the firewall. used tid 3248145 to help with config. everything else in the environment (4 x mta, 1 x gwia, 6 x poa, 3 x webacc) is also running 8.0.0 HP2.

novell web team says its something in webaccess that is spiraling out of control eventually eating up all the memory causing tomcat to crash. how can i make it stop?