We have two node Oracle 9.2.0.4 Enterprise Edition cluster in Red Hat AS
2.1. Both the instances that are in those servers read/write to
OCFS filesystem in SAN. All the hardwares are Dell. Whenever a OS command
that does both io like cp, mv, tar, gzip etc. is run the CPU
usage peaks to around 100%. All the database utilities like exp/imp or
DMS/DLLs run fine. In other words no complaints on database side.
Currently we can't move/copy database backups because of this or we can't
run RMAN in NFS. We have asynch IO turned on, if that has to do
anything with this problem...Any pointer to resolve this problem would be
greatly appreciated.
Thanks,
SP
Received on Fri May 13 2005 - 13:49:21 CDT
