Re: "Checkpoint not complete" message in alert.log file

On Tue, 23 Sep 1997 22:57:28 GMT, Cezariusz.Marek_at_comarch.pl (Cezariusz Marek) wrote:

>Which is better: to add redo log group (or increase the log file
>sizes) or to start the Checkpoint process (what is checkpoint prosess
>and what it gives to me)?

adding the checkpoint process creates a dedicated program that runs with the sole purpose of writing your redo log files. If your database is CPU bound, this might help, because it gives another process the chance to write your log files. If the problem is your disks are too busy to complete a log file checkpoint before the next is ready, this process will just wait too. And you will get the same warning.

You may want to look at moving the log files to less busy drives. Log file buffer size may help. It all depends on your configuration, and performance stats. What is so busy on your system, that the log file writes are waiting? That's the area that must be compensated for. Received on Fri Sep 26 1997 - 00:00:00 CDT