Summary: Can a user-process cause Solaris to boot-up, with having shutdown?

Thanks a lot to all,

Most responses agreed this can be power failure, or much more likely
Oracle RAC evicting the node and forcing it to shutdown "improperly"
to avoid having it "save anything"...

We confirmed from Oracle RAC log files on other nodes that the host
"elli" was indeed evicted, and also found a bug report mentioning
that:

6389053 Oracle RAC defines "uadmin 1 1" as reboot command, trips over
boot archive check after reset

So that would explain the observed system logs.

Thank you all very much.
Mohamed.

On Mon, Mar 15, 2010 at 11:57 PM, Mohamed Lrhazi <lrhazi@gmail.com> wrote:
> My question in other words is: What can explain a system crash where
> the logs go from "usual log lines" to a sudden system boot messages
> like this:
>
> ================
> Mar 15 21:11:55 elli genunix: [ID 540533 kern.notice] ^MSunOS Release
> 5.10 Version Generic_138888-03 64-bit
>
> Mar 15 21:11:55 elli genunix: [ID 172908 kern.notice] Copyright
> 1983-2008 Sun Microsystems, Inc.  All rights reserved.
>
> Mar 15 21:11:55 elli Use is subject to license terms.
>
> ================
>
> That is, no panic message, no "am going down..." or anything at all!
>
> Could it be power failure? Could there be other explanations?
>
> This system is part of an Oracle cluster, and I don't know the
> software at all, but from OS perspective, am tending to think that
> Oracle is just a user process and could not be doing anything behind
> the kernel's back, but maybe I am wrong.
>
> I also don't know with the linefeed character doing there in the first
> boot line!
>
> Thanks a lot.
> Mohamed.
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers