Version details : DB2 V6.1; OS390 V2.8

Hi,
One of my bufferpool is having Negative HIT ratio.

For that BP, there is only one large 32 Partitioned TSpace containing a heavily used table (mainly for online txn) of some million rows.

The reason I observed for negative hit ratio, is that, for the tableSpace allocated to the BP the usage of DYNAMIC PREFETCH (Async i/O) is far more than the no of GETPAGES requests.
I used the formula
HITRATIO = (GETPAGE-(SYNC_RIO + ASYNC_i/o) )/GETPAGE.

We tried with increasing the BPSIZE to three fold ...but didnot show any improvement in performance.

Could anybody help me out how to control the DYNAMIC PREFETCH ??

Thanks in advance !!!
pm.

Is it a problem?

http://publib.boulder.ibm.com/infoce...jndmstr787.htm

Have you looked at omegamon?

Sounds like it's threading out.....