RE: I/O and db_file_multiblock_read_count

 
well, just eyeballing the math...it looks
like 93% less trips into the kernel and 
an approximate 85% reduction in system time.

looks about right to me...


>>>
>>>Strangely, the sys time is much lower with the 1M blocksize. 
>>>Is there an explanation for that?
>>>
>
--
http://www.freelists.org/webpage/oracle-l

Other related posts: