That's not particularly interesting. Linux has maximum atomic I/O read smaller then 1MB (I believe it's 256KB?) so using 1MB blocks will Inevitably break one I/O request into many. _____ From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Jared Still Sent: Monday, December 11, 2006 3:46 PM To: kmoore@xxxxxxxxxxxx Cc: oracle-l Subject: Re: I/O and db_file_multiblock_read_count Interesting that the 64K blocksize significantly outperforms the 1M blocksize.