Re: Oracle 7.3 on HPUX 10.01 - any results yet?

In article <MQgSzAAwKdByEwpJ_at_chin.demon.co.uk>, chin_at_chin.demon.co.uk says...
>
>In article <AE27FFE99668166EA_at_fjp.his.com>, Fred Puhan
 <fjp_at_his.com>
>writes
>>In article <4trlkk$elf_at_vixen.cso.uiuc.edu>,
>>pjackson_at_prairienet.org (Paul C. Jackson) wrote:
>>
>>>Has anyone taken the leap to 7.3 yet? If
>>>so, how is it working? Any bugs?
>>>
>>>We are very apprehensive since our last upgrade
 attempt.
>>>
>>>Any insight would be appreciated.
>>>
>>>Thanks,
>>>
>>>pjackson_at_prairienet.org
>>>--
>>
>>I installed 7.3 on an HP 9000/D250 running HP-UX 10.01
 and am running two
>>instances currently.
>>
>>Oracle forgot to include references to a required shell
 script
>>(STARTUP.SH?) in the installation instructions, but
 other than that,
>>everything has been running smoothly. The most
 noticeable change (for me,
>>at least) is that Server Manager has now officially
 replaced SQL*DBA. As I
>>am not fortunate enough to have an XTerminal, I am only
 able to run Server
>>Manager in line mode, which means I no longer have
 access to any of the
>>monitoring tools.
>>
>>For some reason, I did not receive the CD that
 contained the Windows-based
>>client tools, and so I had to manually create and
 configure the SQL*Net
>>files (tnsnames.ora, sqlnet.ora, listener.ora).
 Despite the written
>>documentation Oracle says they _will_ support such
 configurations (the docs
>>now state that users MUST use the Network Manager tools
 to create these
>>files).
>>
>>I also obtained and installed four patches from HP that
 Oracle recommends.
>>
>>Not a painless installation, but not a showstopper,
 either. The product
>>seems stable and works as advertised.
>>
>>
>>
>>
>We have also had similar experiances. The only
 difference is that we
>found a bug with Oracle (actually a hang over from V7.2
 that was not
>fixed properly). If multiple database writers are used
 and the extends
>option is used with tablespaces then when one db writer
 extends a
>tablespace the others do not know about it and a
 database shutdown and
>media recovery are required (assuming archive logging is
 on) to recover!
>
>For now only run with a single DB writer or do not have
 dynamically
>extending tablespaces.
>
>The biggest cultural differences are not sqldba as
 mentioned (which
>works well with and appropriate X server), and the
 absence of SQL*Net
>V1.
>
>It has been running for 2 months now with little
 difficulty and is a
>little faster than the old V7.1.4 we used before. There
 may be other
>wrinkles awaiting in the new functionally which we have
 not yet had time
>to play with.
>--
>Chris Dipple

Also, Oracle did not change the dbshut and dbstart scripts in ORACLE_HOME/bin directory so they still reference sqldba instead of svrmgrl. The only work around is to link svrmgrl to sqldba with:

ln svrmgrl sqldba

Or change it in the scripts (Oracle does not support this). Received on Wed Aug 07 1996 - 00:00:00 CEST