Here is the crs log. I guess what I was wondering was when Oracle will
reboot a node and when it will shut it down. Documentation seems to be
weak when it comes to describing these scenarios.
Thanks.
Henry
K Gopalakrishnan wrote:
Henry:
What does your CRS log file say? I guess CRS killed instance to avoid data corruption. THis seems like a simple io-fencing issue.
Regards, Gopal
On 11/15/05, Henry Poras <henry@xxxxxxxxxxxxxxx> wrote:
I've been playing around with a RAC system (10.1.0.4 on RedHat AS3), and managed to crash it(no problem, I'm trying to see what can go wrong). It was probably due to a corrupted Registry file (does ocrconfig -import work?). On trying to export and then import a Registry file, then restart CRS, one of our nodes removed itself (no longer visible from the network. Probable crash). When our SysAdm went down to look, he said Oracle had 'halt'ed the server. All the doc I have seen talks about 'reboot' under certain conditions, but I have never seen a halt (or shutdown) mentioned anywhere. Has anyone else seen (or seen documentation of) this behavior?
Henry
-- Best Regards, K Gopalakrishnan Co-Author: Oracle Wait Interface, Oracle Press 2004 http://www.amazon.com/exec/obidos/tg/detail/-/007222729X/
-- //www.freelists.org/webpage/oracle-l