Initiating a quicker checkpoint
Hi,
during peak hour, when user complains about they are getting slow response, we have checked alert log and found messages like "checkpoint not completed", when i query v$log, we always find status ACTIVE for all the redo log group Except the current redo log group.
My redo log group size was 50mb initially, due to the issue i increase the redo log group to 200mb.But after increasing the redo log group,still the problem persist. May be to some point it is normal to show ACTIVE status based on workload.But as we increased the size much bigger than than the old one, the things should be much better.
We are not using FAST_START_MTTR_TARGET, we use log_checkpoint_timeout to control checkpoint which value is set to 1800. Now i have decided to decrease the size to 300 so that checkpoint initiate more quickly. And i think the performance issue related with the redo log groups. Because unless checkpoint is not complete for a redo log group/or it render as inactive,the group can't reuse.
Am i right?? Anyone with better idea would be very helpful...