Redo logs, poor writing performance
groxyNov 12 2007 — edited Nov 12 2007I using Oracle 10g2 on Suse 9 x86_64. The storage is based on RAID5 composed of 3 disks
In my system I have a lot of small transactions. This cause a huge amount of writes into redo logs, which seems to cause the bottleneck of the system and latency on workstations. I would like to improve the performance.The 3 redo log groups are written to two location on the same logical drive:
+DATA1/server/onlinelog
/FlashRecoveryArea/Server
The size of redo log file is 51200 KB
How can I improve the performance without changing the RAID configuration?
Thanks for help
groxy