I have some machines runing Redhat AS 3 update 4 and the OSs are randomly crashing. The only thing that they have in common is the OS. A reboot fixes it but I am looking for a solution. Any ideas or directions to head would be most helpful.

Define "crash".

Box hangs?

Box goes to Panic and hangs there?

No messages on console but it just locks up?

Oddly we just saw issues on our RHEL AS 3 box where it was doing a Kernel Panic then hanging there. We'd have to hard boot it because Ctrl-Alt-Del had no effect. Our current theory is this had to do with an update to the Java under Oracle the DBAs did (to 1.4.2.-12) for DST. At the OS level we still have 1.4.2-08. Under Oracle they had both 1.4.2-08 and 1.4.2-02 previously. The Panic occurred when they used the 1.4.2-12 initially so they went back to the earlier version. However time was off so they went back to 1.4.2-12 and it Paniced a second time. We chose to run the tzdupdater (Java DST fix) from Sun agains the earlier 1.4.2-08 and 1.4.2-02 and have been running on that for a couple of days with no crash.

I'd also seen a lockup previously on a system when there was a hardware fault. In that situation we got messages from Megaraid (the driver used for DELL [LSI] PERC RAID cards) on the console and the system just hung. We ended up replacing the PERC card.

In all events the power cycle of the server got it going again.