s.kapitza wrote:
> hello,
>
> puritycontr0L_at_netscape.net (Purity Control) wrote in message > As an experiment, I copied the balky database to another NT 4.0 SP6
>
>>Oracle 7.3.4. server and recreated the services, rebooted. Exact same
>>behavior with the services. I've never encountered such a balky database.
>
>
> wouldn'd this imply that the problem is in the database/init.ora ?
> - any difference in the init.ora (passwordfile, os auth ... ?)
>
> Does a newly created DB show the same behaviour ? If not i would
> recreate the balky db or import it in the newly created.
>
> I must confess that haven'd understand this "connect internal" thing
> fully
> but - you wrote that you have to enter the password when connecting.
> Have you try'd to sync the internal passwd with sys user ? It seems
> you have to enter the password for internal.
>
> Are you sure that the instance is up, when the server boot's ?
> as you wrote :
>
>
>>SVRMGR> connect internal
>>Password:
>>Connected to an idle instance.
>>SVRMGR> startup
>>ORACLE instance started.
>
>
> it doesn't look like.
>
> regards
>
> s.kapitza
>
>
> news:<dc90049d.0407231026.729ca6bc_at_posting.google.com>...
>
>>Frank van Bortel <fvanbortel_at_netscape.net> wrote in message news:<cdnurf$m2j$1_at_news2.tilbu1.nb.home.nl>...
>>
>>>Purity Control wrote:
>>>
>>>>Frank van Bortel <fvanbortel_at_netscape.net> wrote in message news:<cdldb8$fdd$1_at_news1.tilbu1.nb.home.nl>...
>>
>>[liberal snippage]
>>
>>
>>>>The behavior I am seeing is that both services go to a status of "Started"
>>>>and OracleStartSID never completes. This is why the background services are
>>>>in memory. It cannot complete, because the service is unable to mount
>>>>the database files. This is the real mystery -- what can't it mount
>>>>the database files?
>>>
>>>Where do you have that information from?!?
>>
>>Logical conclusion from the Alert Log and Task Manager. The Alert Log
>>clearly shows that all the background processes have started, and the
>>Task Manager shows the SGA in memory. The next process in a startup
>>is to mount the datafiles and open the database. The datafiles never
>>mount.
>>
>>
>>>I've seen similar behaviour
>>>on the same platform with clusters, where the service simply was faster
>>>than the cluster s/w. The drive(s) on which the data files resided,
>>>simply weren't available at the time the service was started.
>>>- move the data files to direct attached storage devices (and
>>> prove the point)
>>>- make the service dependant on the cluster services.
>>
>>This is a Windows NT PC, not a unix box. The disks are available at
>>boot. The fact that the server has 2 databases on it, and the
>>datafiles for both are on the same drive (F:) is more proof
>>that the device F: is available. One database opens just fine,
>>the other balks at the datafile mount on reboots.
>>
>>
>>>Hm - symptoms, that would subscribe to the above.
>>>What about a Command Line Interface (CLI - or "DOS Box"), and:
>>>net start OracleStartVAPP
>>>or
>>>net stop OracleStartVAPP
>>
>>Yeah, done that many a time. It starts and stops
>>cleanly from a Command Prompt window, just not
>>at a server reboot.
>>
>>I really need this database to open at reboot. I'm shipping
>>this server to a site with no DBA. All they are going to know
>>how to do is reboot the server. If the database doesn't open
>>at reboot, they'll be lost.
>>
>>
>>>What about the NT logs (specifically: Application and Security
>>>Logs) - any failure codes there?
>>
>>Event logs are clean.
>>
>>Thanks.
One thing strikes as odd: logged on with the
correct credentials, I don't believe you ever
have to supply a password, when logging on as
internal. The fact you are being prompted for
one, could point to the direction of being
logged on as something else than local admin.
--
Regards,
Frank van Bortel
Received on Sat Jul 24 2004 - 08:29:32 CDT
