Hello,
I've noticed that some of our SNMP sensors repeatedly give no response errors/warnings that I believe to be false negatives, and I was wondering if anyone had any ideas to alleviate this, as I feel these specific monitors aren't useful in any sense in their current state. I will note that these are Linux boxes that are giving the errors, and our Windows servers/clients have not had similar issues. It also seems to be the same devices every time. I'll attach some screenshots of our 2 hour graph to show just how intermittent it is.
Some things we have tried:
Shuffling devices to new remote probes for load balancing
Increasing the SNMP timeout to various amounts (I've tried anywhere from 5-60 seconds, but the devices fail to respond for minutes at a time, so I'd have to set realistically a 15 minute timeout to avoid 90% of these warnings and errors, which is far from ideal).
Any advice is greatly appreciated!
Article Comments
I have a similar issue. I am starting to get these same SNMP errors and "dropouts" on a Sonicwall TZ300 firewall that has been working fine for over 2 months. I have a second identical firewall that is not having these issues. The main difference is that the "failing" firewall is on the other end of a VPN tunnel.
My PRTG host is a Windows 2012R2 VM.
Jun, 2017 - Permalink
Hello trex1000,
Please find here what to do about intermittent SNMP dropouts on Sonicwall firewalls.
Kind regards,
Erhard
Jun, 2017 - Permalink
Just another potential solution
Using the PRTG cloud service: The probe was on a device using the FQDN in the LAN, rather than IP.
But there were two DNS entries for this server, so the SNMP, source address kept "switching", we had to permit SNMP from both IPs and all was good.
Apr, 2023 - Permalink
Hello there,
Everything you describe points to an issue with the SNMP implementation on the affected devices, especially the SNMP daemon seems to be "shaky" I'm afraid.
The sensors do the same job as on other devices where you don't experience these issues, so it does not appear to be a general issue in the network or with PRTG so far. Raising scanning intervals is a good approach with such "shaky" devices, also you can try making use of "SNMP Compatibility Options" located in the device's "Settings" tab in PRTG. Try setting an SNMP Delay and also give "Use Single Get" a shot. Another setting worth fiddling with is "SNMP Timeout" located in section "SNMP Credentials", also located in the device's "Settings" tab. Other than that you might find other resources pointing to what could be improved around the target host's SNMP configuration.
See also here our checklist for setting up SNMP on Linux.
Kind regards,
Erhard
Feb, 2017 - Permalink