APC UPS - Basic Node Information Does Not Update

Good morning!

I've noticed that several of our APC UPS nodes in Solarwinds are showing old data (serial numbers, software versions, etc). The UPS' in question have been replaced, it looks like Solarwinds is still showing the data from the old UPS instead of the new. I've tried rediscovering the nodes and disabling/re-enabling the UPS poller under list resources. Short of nuking the node(s) and recreating them, are there any suggestions what can be done to get these nodes updated?

For the record, we're on NPM 2020.2.6.

Find more posts tagged with

npm

apc

ups

Accepted answers

All comments

superfly

Do you have NCM? If so, try running an "Update Inventory" job on these UPS's.

Fred.Loucks

Is the new UPS unit the same model number as the old UPS unit?

svindle

This did not help.

svindle

Similar. APC Smart-UPS X 2200 is the old, APC Smart-UPS 2200 is the new.

Fred.Loucks

Similar or same? APC Smart-UPS X 2200 are newer models. APC Smart-UPS 2200 are a model made years ago.

1) OIDs differ in some "X series" smart-ups to the older Standard "Smart-ups series" ups units. You might have an OID issue. I've ran into a similar issue on some devices that were swapped out. Stale information was displayed because new information couldn't be pulled. There would be a note saying last successful poll was X amount of time ago.

2) APC has various NICs for the Smart-UPS X 2200s. APC says they have different manufacturers that make those NICs. APC listed different MIBs for various devices and parts/NICs on their devices. It would take awhile, but you might what to check your OIDs vs what APC published. The NIC might not be the same as the NIC in the unit you removed.

Hope this info helped. When you find the solution, can you post it for us? Seeing different fix actions helps me out and gives me better ideas in the future.

nickburns186

This happened to us after the latest update to 2022.3.0. Had to run a manual rediscover on all APC UPS's in our environment to get polling back but otherwise fine so far.

svindle

Agree, similar is not the same; However, I'm not using custom UnDP's for this, I'm using the in-built UPS poller.

It works if I delete an affected node and then add it back, but then I have to rebuild dependencies and fill out custom properties, and doing that en-masse is prohibitively time consuming.

Martin_Crothers

Hi @svindle,

I was under the impression that when a node is 'physically' swapped out (hardware replaced, VM recreated, etc.) that it should be deleted from Orion and recreated as a new node.

Could be that there's actually a way to get around doing that, but I'm not aware of any method(s). So, delete and rediscover is how I handle this situation.

nickburns186

If I'm swapping out exact hardware, then a simple rediscover is usually all thats needed. No need to delete first.