Interference of snmp_exporter processes

[Andy1616]

Hi out there,

I do run several instances of snmp_exporter in same VM, for different HW vendors and for different schedules.
So e.g. I have a instance for Cisco/30sec, another one for Cisco/60sec, one for Dell/30sec, one for Dell/60sec

Having had only a/30 and b/30 all was fine.
Enhancing and also starting a/60 and b/60 I see unexpected impact on a/30 and b/30 like:

• exporter metrics are getting slower in response
• number of fds increasing
• number of go routines increasing

How to the instances affect each other?
Did not expect any side effects like that...

Any idea / feedback is welcome

[Andy1616]

well - found it - all of the instances have been served by a single apache for caching which in the end became the bottleneck..
DONE