My guess will be that PHYRDS is the count of start i/o's. Each start i/o
read mutilblock_read_count blocks from the disk.
The data buffer that you read with each start i/o is the same: 8 blocks of
8k or 16 blocks of 4k. So you get the same number of start i/o's.
Yechiel Adar
Mehish
- Original
list,
i'm doing benchmarking using two DB's with different block size
i run a count(*) on a 17 million row table, and compare the sql_trace file
and the v$filestat stats..
the db was bounced before each test, the init.ora params were identical,
EXCEPT
in DB1 (4k block size) the muldiblock read
