Hi All,
Recently I have hit a bug "Bug 10194190 - Solaris: Process spin and/or ASM and DB crash if RAC instance up for > 248 days ( Doc ID 10194190.8 )"
A workaround is to regular reboot the DB server.
I would like to know whether reguler rebooting a DB server is a good practice in a normal case without hitting the bug.
In my experience, Unix platform is more stable than Windows platform in
term of up time. Undocumented truth that we need to restart a Windows
server to get services resumed in some cases.
I have not heard of restarting a database regularly in Unix platform.
In my knowledge, a critical 7x24 services should have as longer uptime as possible.
However, I do agree a database should be rebooted in some cases like
patching and some other maintenance.
In our case, I agree that we need to take this workaround, regular reboot the DB server, to
resolve the problem at this stage. However, I want to know, as a good practice, whether a DB server should be restarted regularly in the
normal case without hitting the bug.