I've to monitor about 140 Cisco Catalyst Switches.
From time to time I'd like to monitor all the ports on all Switches for Input and output errors.
That are about
140 Switches * 26 Ports * 2 Tests (IN- and OUT-errors) =
7000 tests to perform.
I wrote a script to create a txt-File to use for import to HM.
Switch1 SNMP InErrors If1
Switch1 SNMP InErrors If2
.
Switch1 SNMP InErrors If25
---
Switch1 SNMP OutErrors If1
Switch1 SNMP OutErrors If2
.
Switch1 SNMP OutErrors If25
------
Switch2 SNMP InErrors If1
Switch2 SNMP InErrors If2
.
Switch2 SNMP InErrors If25
---
Switch2 SNMP OutErrors If1
Switch2 SNMP OutErrors If2
.
Switch2 SNMP OutErrors If25
---
and so on ...
On the one hand I'd like to set parallelism as high as possible ("Don't start more than 128 tests per second")
When I do that every switch is fired with 2 * 25 snmp requests and collapses

What I need is:
- Don't start more than xxx tests per IP-Address (host) per second
- What influence do I have on the sequence of the test ?
- Possible solution would be to Test
Switch1 SNMP InErrors If1
Switch2 SNMP InErrors If1
.
Switch140 SNMP InErrors If1
Switch1 SNMP InErrors If2
Switch2 SNMP InErrors If2
.
but how to force that ?
Ohter suggestions are welcome.
Sorry for that long description, hope the problem is clear.
Juergen