Hi folks,
I'm using rdbms EE 11.2.0.2 on Solaris 10.
Yesterday, we had a very high usage on our top activities graph in OEM.
I've run addm and one of the recommendations it says is: "investigate improving the performance of I/O to the online redo log files."
So, what kind of things can be done to improve the performance of the I/O on the online redo logs?
They are already on their own LUN that is striped across 90 drives in our SAN.
The log sizes are sized such that they switch approx. every 45 minutes to an hour.
What else can I consider to do to improve their I/O performance?
We are running the database on an older DL360 G6, with dual quad-core CPUs, using fiber to connect from the HBA cards to the SAN.