We have a database version 10.2.0.5, patch 14 running on a windows server 2003 (x64 bit). The server has 32G of ram. The swap space is 1.5 x memory.
The database instance has an SGA_TARGET=10G, PGA=2G. Every now and then, the entire server reboots automatically. This happens a least once a week for the last year. The is nothing in the windows event logs that shows any sort of problem. The only thing I see is in the alert log of any relevance is 'ORACLE V10.2.0.5.0 - 64bit Production vsnsta=0 vsnsql=14 vsnxtr=3' during startup. Any idea why this might be happening ? I opened a ticket in metalink and was asked to install the latest patches. The problem is still occuring. Anybody experienced the same problem ?
Disclaimer: Advice is provided to the best of my knowledge but no implicit or explicit warranties are provided. Since the advisor explicitly encourages testing any and all suggestions on a test non-production environment advisor should not held liable or responsible for any actions taken based on the given advice.
It could certainly be a Windows issue, however it is worth pointing out the PGA memory is not fixed and it c an grow, so if there is a process or processes that are using up 20-22GB of RAM, once the virtual memory is exhausted, its possible that Windows could reboot. You might want to look for memory intensive processing, list PL/SQL collections to see if there is something that is using up all of the memory on the server.
Barring that you should look at Windows and any 3rd party device drivers that you might have loaded.