archive_lag_target tips Server Tips by Donald BurlesonFebruary 5, 2015

Question:  I have implemented Data Guard successfully with the help of the people on this website! Thanks. I have set the archive_lag_target = 600 (10 minutes) and everything seems fine, except I get the follow error on the data guard screen in Grid Control.

ORA-16792: configuration property value is inconsistent with database setting

It seems to function fine, but if I press the 'Verify Configuration' button in Grid Control on the Data Guard screen, it goes though the verification process, sets the archive_lag_target to zero, and clears the ORA-16792 error. I set the value for archove_lag_target like this.

alter system set archive_lag_target = 600;

I tried using this command

alter system set archive_lag_target = 600 scope=spfile;

and the message said 'system altered' but the value for archive_lag_target was still zero.  The value also sets itself back to zero during a shutdown/restart when I do nightly backups.  What am I doing wrong?  I used the alter system set archive_lag_target = 600 scope=both; and it sets the value correctly, but it still sets things back to Zero during a shutdown/startup and I'm still getting the ORA-16792 error.

Answer:  On the data guard verification wizard in Grid control I found a Broker Managed option and a DB option for fixing the problem. I'd been using the Broker option, so I used the other one which I believe wrote the value to the init.ora file, and this seems to have fixed the problem.

archive_lag_target for RAC

In RAC, the online redo logs should be sized to see three or four log switches per hour at peak DML load. In the sample output above, there was at least one hour in one instance that experienced eight log switches.

It is a best practice to have all redo logs sized the same for each thread and for each group in that thread.
 
Sizing the redo logs to meet the target of three or four log switches works well for the hour that has the highest redo generation rate. But what about the hour timeframe that has a very low redo generation rate?

Some database administrators translate the three or four log switches per hour to a requirement that the redo logs should switch every fifteen or twenty minutes. As such, they often set the archive_lag_target parameter to 900 or 1200.

If you like Oracle tuning, you might enjoy my book "Oracle Tuning: The Definitive Reference", with 950 pages of tuning tips and scripts.  You can buy it direct from the publisher for 30%-off and get instant access to the code depot of Oracle tuning scripts.