>>>I do recall the umpteen times folks reported here and >>>elsewhere no change or incoherent results in their >>>cooked/dio/aio testing and the problem was, in almost all >>>cases, OS and/or Oracle incorrectly or incompletely patched! Nail on the head, Nuno. But the fact that anything has to be patched to enable async IO is what has always seemed completely rediculous to me. This is, historically, the reason Oracle proposed the Oracle Disk Manager library specification so that async IO would be implemented the same on every platform so Oracle porting doesn't have to munge through heinz 57 variety of IO apis to get the OSDs right.. ODM: http://www.oracle.com/technology/deploy/availability/pdf/odm_wp.pdf -- //www.freelists.org/webpage/oracle-l