what actually happens during...

Hi,
I'm trying to optimize some DB2 queries and I'm taking benchmarks with db2batch.

I see timing for execute and fetch is reported in the output.

I'd like to understand what DB2 does during the execute, and what during the fetch.

For example, where is the I/O time (time spent loading data from disk into bufferpools) counted?

What can be done to enhance fetch rate, given that CPU cannot be changed?

How are these times in relationship with the total cost of the access plan?

Thanks