I've mentioned this once before myself. I have a *maximum* window of about
14 hours, once a week, that ADSM can be down. This is about to shrink to 12
hours. If I get a corrupted database again (it's happened once), I'll open a
PMR and follow the suggestions for a fix. If the fix (auditdb) doesn't
complete in the available window, I'll be opening a second PMR on the
inadequate repair utility.

ADSM should be using a *real* dbms, whether DB2, oracle, Informix, or
something else. I'd prefer interchangeability here -- I've two trained
Oracle DBAs who tune, maintain, and (if needed) repair our oracle databases.
It would be real nice for the ADSM database to be 'just another database
instance'.

Tom Kauffman
NIBCO, Inc.

> -----Original Message-----
> From: Richard Sims [SMTP:rbs AT BU DOT EDU]
> Sent: Thursday, April 22, 1999 6:38 AM
> To:   ADSM-L AT VM.MARIST DOT EDU
> Subject:      Re: Strange Error Deleting a Disk Volume
>
> >IBM has suggested that probably the database is corrupt and that we
> >should do an AUDIT DB FIX=YES.  Sigh............
> >We're going to do that this weekend.
>
> It will likely mean having your server unavailable for most of the
> weekend.  The ADSM product is long overdue for online auditing of
> the database - it should not be necessary to take the whole facility
> down for days to detect and correct an enduring inconsistency in a
> small part of the database.  That's ridiculous.  ADSM should be a
> far more sophisticated product than that.
>      Richard Sims, BU