My thanks to Raul Sossa, Alex Harkema, Cathy Coleman, and Wade Buchanan for
responding with their suggestions. However, in the end, it was Compaq
Support that provided the solution. In the ASE stop scripts i was using
fuser to clear the mount points prior to running the Oracle stop scripts.
When i edited the stop script and moved the fuser commands below the Oracle
stop scripts, it worked like a champ. As this is an inherited system, i
can't comment on why things were like they were. However, i can add that
this problem did not manifest itself until we upgraded from Oracle 7.3.4 to
Oracle 8.1.6.
Thanks to all.
> -----Original Message-----
> From: Scott Hyjek
> Sent: Friday, May 11, 2001 12:20 PM
> To: Tru64 UnixMgrs (E-mail)
> Subject: Oracle 8.1.6 + ASE v1.6 + Tru64 v4.0F...problem with disk
> service stop scripts
>
> We've got a problem within our ASE stop scripts for disk services.
> Specifically, we try a relocate and it fails after the execution timeout
> expires. Because the disk services hold Oracle 8.1.6 databases we unalias
> the database instance from tu0, then clear the mount point with fuser,
> then execute a start/stop script from /sbin/init.d...for example
> /sbin/init.d/oracle_lxor. that script calls (via su - oracle -c), in turn
> a specific instance start or stop script that sets a few variable and then
> executes "svrmgrl". I can provide the scripts, if necessary.
>
> Our current theory suggest that the ASE stop script is "sticking" within
> the second script called at the "svrmgrl" line. Anybody run into this?
>
> shyjek_at_rbmg.com
>
>
> Scott Hyjek
> "Little Man...Big Attitude"
> Network Engineering
> 803-741-3101
>
>
>
Received on Fri May 18 2001 - 14:13:49 NZST