weird snmp voltage issue

[cherwilco]

WS-8-150-DC with 1.55 FW. I am using this oid for input voltage in PRTG: 1.3.6.1.4.1.46242.4.1.3.4

I currently have this switch in our test tower outside my office running off a Simpliphi 48v lithium battery charging via solar.

when I login to the switch GUI I am seeing the input voltage as 52.9, and when I meter the input legs on the switch I see the same voltage. In PRTG with that oid I am getting 53.9 with a fresh scan.

I also have another of the same model switch on my test bench on the same firmware being fed 52.9v off of a bench power supply. In PRTG on this switch I am reading 53.9 with a fresh scan.

with that same bench test switch I set my bench supply to 24.7V and in PRTG its reading 25.6V

I then set it to 13.5V and in PRTG its reading 14.5V



I know that 1 volt difference isn't too important with a 48 volt battery but I have a bunch of these same switch's deployed on 12v lithium battery's and 1 volt there makes a big difference especially when using low voltage shutdown on the poe ports.

am I using the correct OID?


EDIT: as my bench testing switch sits with the bench power supply at 13.5 I noticed my PRTG scan just read 13.5.... until the next scan when it went back to 14.5 subsequent scans after result in almost always 14.5 with the occasional 13.5??

[Stephen]

I'm reconfiguring my lab to check this (I was hoping to have it done before I responded, but other responsibilities got in the way)

But it is possible there is a bug in SNMP causing this so I will try to replicate and patch it. However, I don't think you'll have to worry about issue's with watchdogs or low voltage shutdown because they are watching the same values that you see in the web UI, not SNMP. So theoretically they should behave correctly.

Still, if it is a bug in SNMP I will track it down.

[cherwilco]

I have an open ticket with Praessler. I'll let you know how that goes. Early this morning I DID notice that the graph in PRTG is showing correct voltage values for both of my test setup, the tree view in prtg is showing the incorrect voltage.... most of the time (subsequent refresh's do get it to occasionally show correct voltage)

so this could be a bug in prtg unless the graph only updating every 15 mins just so happens to be missing the times when it jumps up 1 volt... IDK yet

good point on low volt shutdown not being affected by this, looks like so far this is really only interfering with our logging and not operation

[cherwilco]

Update: Praessler replied, it is a confirmed bug on their end. Hope you left your lab the way it was Stephen!

[Stephen]

Thank you for the update cherwilco,
No worries, my lab needed some redecorating anyway.