Subject: | |
From: | |
Reply To: | |
Date: | Fri, 29 Jan 1999 02:57:54 GMT |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Hi Joe,
I had a similar problem last night while changing a fiber card in our
9000. I know its not a 3000, but maybe this will help. The listener
was shutdown while processes were still connected and Oracle was stuck
in limbo. I couldn't shut it down and I couldn't start it up. Well I
was already at a shutdown so what to I do ... next step was to flip the
switch. I'm sure the database will come back fine ... yeah right! A
quick call to the D.B.A. suggested a shutdown abort. I got lucky, it
worked and I was able to contniue.
Again I don't know if this would work on the 3000, but give it a try
next time your in that situation.
Joseph Rosenblatt wrote:
>
> Last night I had to bring down my network. I did this with the standard
> NETCONTROL STOP NSCONTROL STOP commands. To my amazement this caused ORACLE
> to come down. I guess I can understand why this happened, something like
> disruption of the socket but the story continues. The standard recovery
> procedures start with stopping ORACLE. The reply was can't stop ORACLE it's
> not up. Go to bring up ORACLE it tells me its already attached. Can't go UP,
> can't go DOWN. It's time for a support call. The gist of the response was
> ORACLE is hung in limbo, Reboot. I asked," Use START RECOVERY or START
> NORECOVERY?" The reply was, "Huh?" I did a START RECOVERY (gotta save those
> user streamed jobs in the queue) and was able to restart ORACLE.
>
> My suggestion for an ad slogan: ORACLE! WHEN DOES THE FUN START?
>
> Just my thoughts.
|
|
|