Re: Help!!! ORA-04031

From: Paul Brewer <paulb_at_pbrewer.demon.co.uk>
Date: 1996/10/14
Message-ID: <pXLRHHAnbrYyEwk0_at_pbrewer.demon.co.uk>#1/1


In article <32627EBB.BDD_at_mailport.delta-air.com>, Bill Roney <bill.roney_at_mailport.delta-air.com> writes
>I'm working on a project that has now run across this error in two different
>processes. We have
>done as the Oracle books suggest and bumped up the parameters in the init.ora,
>but still get it ...
>we just delay the frequency.
>
>This message seems to occur after a large number of committed transactions.
>Shouldn't the shared
>memory be deallocated and reclaimed by Oracle? The transactions are each
>independent connections
>scattered throughout the day ... the latest cases occurs after approximately
>170,000 transactions.
>Is there some sort of magical, "this is for OLTP" switch somewhere?
>
>Thanks in advance for your assistance.
>
>Bill Roney
Been having the same problem, and lurking here. Apparently, this is a bug; the freed memory isn't being reclaimed properly, so the shared pool just fills up over time. I'm no expert, but I read that the workaround is to run 'alter database flush shared_pool' periodically, until you get the relevant upgrade.

PS Feel free to get a second opinion - I'm a novice just trying to help.

Paul Brewer Received on Mon Oct 14 1996 - 00:00:00 CEST

Original text of this message