Re: Mysterious log switches

From: Jeff Shervey <jeff.shervey_at_us.landisstaefa.com_spamblock>
Date: 1998/02/16
Message-ID: <6ca66u$raa@news.us.landisstaefa.com>#1/1

You don't by any chance have the *init.ora param: "log_checkpoint_timeout" set to about 36,000 seconds? Check log_checkpoint_interval. Setting it higher than the number of OS blocks (redo logfile total size) found in your REDO logfile causes it to do a logswitch only when the redo logfile fills up. Setting it smaller speeds recovery and causes logswitches more frequently based on when "log_checkpoint_interval" number of blocks are written to the redo file. There also is a "alter" command that can force a logswitch.

Jeff Shervey - Landis & Staefa

Van Anh Thi Nguyen wrote in message <34E89547.67D03CAE_at_PO-Box.McGill.CA>...
>Are you sure there are no jobs running?
>
>John Bagley wrote:
>
>> I've got a 7.3 Oracle instance hosting Oracle Applications (Gov't
>> Financials) and running in archivelog mode. I know for certain that
>> it
>> is not being used at all (no user activity and no concurrent
>> requests).
>> However, it has been doing log switches at somewhat regular intervals
>> (about every 10-12 hours) and archiving the redo logs. The size of my
>>
>> redo logs is 4MB. The folks at support tell me that this shouldn't be
>>
>> happening, but no one can tell me why it's happening. Anyone got any
>> ideas/suggestions?
>
>
>
Received on Mon Feb 16 1998 - 00:00:00 CST