How to know which processes are major culprit in the event of network congestion?

[Zhang Zhao gui]

I am using GlancePlus and PerfView on HP-UX 10.20.
I want to know how I can identify which processes are major culprit in the
event of network congestion.
Anybody know which process-related metric can be used
to do that?

[Dan Hull]

There are some good white papers on using Glance available at the following
location:

[Emil Velez]

Check out the PROC_LAN_WAIT_PCT or TIME

this will tell you if your process is waiting for I/Os to complete. There is not really a metric to collect number of Lan I/Os per process.

Good luck