Knowing users/tasks with excesive demand of disk server (read access)

[Solomon]

Using Performance tool and "Avg. Disk Read Queue Length" counter we can
monitor read access for one of our disks server (W2K Server).
Constantly the line is on the top of graphic and we suppose its behavior
should be peaks.
Theirs values are:
Last=59.577 Average=36.981 Min=17.565 Max=148.386 Duration=1:40

Our server performance is very slow when this counter is on top.

How can we kow users or tasks are demanding excesive read access of disk
server?

How much should be normal values to measure the excesive use?

Some corrrective or preventive actions?

THANKS!

 

[Gerry Voras]

Your queue lengths are way too high.