RE: Snapshot (Materialized View) Maladies

  • From: "Bobak, Mark" <Mark.Bobak@xxxxxxxxxxxxxxx>
  • To: <oracle-l@xxxxxxxxxxxxx>
  • Date: Fri, 4 Jun 2004 17:21:09 -0400

There's another one, Jared.  5pt font.

-----Original Message-----
From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx]On 
Behalf Of Jared.Still@xxxxxxxxxxx
Sent: Friday, June 04, 2004 5:08 PM
To: oracle-l@xxxxxxxxxxxxx
Subject: Re: Snapshot (Materialized View) Maladies




> Our main issue with them is that there is nothing reported in the alert
> log and no easy detection method for problem determination when they are
> NOT WORKING!  You must check DBA_JOBS and numerous other tables via
> scripts for exceptions (such as - job running too long, job hasn't run
> at all, job not running at frequency or time specified, etc).  Worst of
> all, sometimes even these queries fail to detect that the snapshot has
> not refreshed, and ain't ever gonna.  In extreme cases, we've had to
> rebuild several million row mviews at the recommendation of Oracle,
> simply because we had a network outage or some other glitch.  
> 
> Any comments or similar experiences?  
> 

Many times.  Network glitches can cause the MV slave/master sync 
to go awry.  I've found the result of it in the mv logs, and could 
determine why they would not refresh, but could not determine how 
it happened ( I forget exact details at the moment ) 

The last time it happened was 2 hours before I had to get to the 
airport to catch a plane for HOTSOS.  :( 

That time, dropping/recreating the refresh group fixed the problem. 

Jared 


Other related posts: