Slightly Odd issue with SNMP and Remote Edgecore Devices

Hi Everyone,

I have noticed an odd issue affecting some of my remote (International based) VYOS devices - All the associated devices are Edgecore SAF51015I

I have a PRTG Server located in Australia which handles monitoring for these devices

But sporadically I am seeing SNMP errors though the Ping sensors and connectivity stay up during this I am not seeing any errors in Var/Log/messages after looking them over but PRTG report -2003 errors or a no response

I don’t believe the issue to be Vyos itself more than likely but wondered if there were any other areas I could look at to determine what is occuring on these devices / if there were any reported issues using these

2 Devices are running the 1.2.4 Edgecore release one of the other sites has 1.2.5 Edgecore on it

I would be happy to answer any questions related to this or do additional tests if I can

Thanks

Hello @JaimeSharp, can you tell what RTT (round time trip) from PRTG to some of the VyOS router, e.g. ping response time?
It will be interesting to reproduce and check this in our infrastructure. I think you understand that SNMP is not a very secure protocol, I propose to use encrypted tunnel between PRTG and VyOS

Hi Dmitry , Thank you for the reply

The RTT for this from the server running PRTG is between 270 to 300ms to our device in London - this was gathered running a ping on the server itself and also testing from the router pinging back to the public IP of the management server

though the trace route from that machine once through all the hops is approximately a total of 5 seconds due to the routing it takes / international links as well as between the different states / countries as expected

An encrypted tunnel certainly would make sense to add for these devices though it is being tested on our secondary server for the moment - the polling interval is set to I believe 10 seconds on these sensors we may need to raise them to a higher value

Given what I am seeing so far I am curious if extracting the MIBS for these devices and recreating the sensors might be worth doing to see if the issue disappears

I appreciate the prompt reply and would be curious to see what your testing yields - I am happy to answer any other questions you might have or review any logs

Thanks

Hello @JaimeSharp, sorry for the delay with my reply

Can you try to increase in PRTG credentials (SNMP timeout) and also trying to change SNMP delay in (SNMP compatibility options)?