Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> ORA-04031 errors

ORA-04031 errors

From: Kenneth C Stahl <BluesSax_at_Unforgettable.com>
Date: Tue, 24 Aug 1999 14:37:32 -0400
Message-ID: <37C2E66B.4FCE6D74@Unforgettable.com>


My users called me today and said that they were experiencing ORA-04031 errors. As I was looking at the situation my first thought was, well, I can increase the setting for the SHARED_POOL_SIZE in my initxxx.ora file. That will certainly help the situation in the long run, but in the short run they simply didn't want to bounce the database right now. So, I decided to perform a "ALTER SYSTEM FLUSH SHARED_POOL". I'm still waiting to hear whether this solved the problems for the users (I provide DBA support remotely, so I don't actually see the users).

This is the first time I have ever supported databases which are brought up and left up unless there is a need to shut things down due to an uncontrollable event (power failure, etc.) because the databases are used 24x7 and in the past I have been more accustomed to systems which are bounced at least once a week.

So, what I'd like to hear from those of you that may have encountered this in the past is if you have any thoughts on just running a periodic cron job on a weekly basis to perform an "ALTER SYSTEM FLUSH SHARED_POOL" regardless of whether the problem is encountered. My suspicion is that while this may sound like a horrible idea, the performance hit that will occur is actually pretty minimal since all of the users are using exactly the same Windows Client application and I can schedule the job for a period of low system activity.

As an alternative, is there anyone who can provide a convincing argument that I should just increase the shared pool size until the errors no longer occur. This would have to be an iterative solution and would take a long time because the new setting would not be picked up for potentially weeks or months and in the interim I'd still have to do the flush. The problem I see with this is that while it may allow me to run for longer and longer periods there is no way I can know in an absolute way that I've finally arrived at the best setting because my next period of continuous on-line usage may be longer than the one just completed.

Ken Received on Tue Aug 24 1999 - 13:37:32 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US