Dear all:
We have a single 11.2.0.3 112 GB instance running with ARCHIVELOG mode active and a FRA of 600GB.
Our Desired Mean Time To Recover is 5 minutes and our current Estimated Mean Time To Recover is 11 seconds.
Flashback database is not enabled.
We have identified as much as 147 redo log switches in a single hour during peak times. We have run the advisor an we are asked to enlarge the redo log file from 300 MB to close to 3 GB.
So far so good.
Unfortunately we have a new set of MV views that can only refreshed using the COMPLETE parameter and since the implementation of those we are having huge problems in the FRA being completely consumed by the archived redo log files. So much that we had to reduce our backup settings policy from 28 to 9 days to leave space of the redo logs. We have also enlarged the Fast Recovery Area Location size, but since the MV refresh times are random, we are struggling with FRA free space.
We have a couple of ideas to fix the MV method, so that either the refreshes do not write redo logs or they are incremental so the impact is not as large.
Parallel to this we have the issue with high number of redo log file switches on peak times.
I wanted to know if enlarging the redo log file sizes to the recommended size would have an impact on the archived redo logs using FRA disk space.
Would enlarging the redo log files also produce larger disk space used for archived redo logs or it would have no impact?
So far the only change we implemented was setting the initialization parameter LOG_CHECKPOINT_TIMEOUT to 0, which had a non-zero value. Log_checkpoint_interval is 0 and log_checkpoints_to_alert is falso.
Thanks and regards,
Roman.