Re: RAC platform usage
Date: Mon, 5 May 2008 08:59:15 -0500
I did check all the logs and even created a case with oracle but couldn't find anything. coincidently, 5 or 6 hours before the hang, one of the 2 nodes rebooted by the cluster because it timed out communication with 1 of the 3 multiplexed votingdisks. oracle couldn't give an explannation why the node rebooted even though there were 2 other votingdisks available. oracle suggested to run OS Watcher to wait until it happens again but it hasn't happened yet.
As for the "I am rebooting" message, I forgot the exact word, but when you ssh to the server and run reboot, it is the information it tells you.
On Sat, May 3, 2008 at 7:12 PM, Yong Huang <yong321_at_yahoo.com> wrote:
> Hi Li, did you go through a very extensive check and read of all possible log
> files? I did, for our server hangs. If absolutely nothing is found, including
> checking server consoles, talk to the storage team. Sometimes when they do
> something, it could cause a mysterious hang. Don Seiler's case in September
> last year is very representative of the cases I know
> (http://www.freelists.org/archives/oracle-l/09-2007/threads.html). If that's
> not the cause, then I'll call it truly mysterious.
> Where did you see "I am rebooting"? Console? Get the exact text and we can
> search on Google.
> Yong Huang
> --- Li Li <litanli_at_gmail.com> wrote:
> > we had 1 incident of mysterious hang on a 2-node RAC on linux x86-64.
> > We had another interesting incident on a 5-node RAC on linux x86-64,
> > at one scheduled maintenance, 3 of the 5 nodes wouldn't reboot. we
> > tried "reboot", "shutdown -h now", "init 6", etc, it showed that "I am
> > rebooting (something like that, don't remember exact word)" but it
> > actually didn't reboot. We had to power recycle them.
> > All of these servers are Dell PE 2950, wonder if anybody had the same
> > experience? :)
> Be a better friend, newshound, and
> know-it-all with Yahoo! Mobile. Try it now. http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ