Network Discovery of Already Managed Nodes

From my experience there shouldn't be anything to worry about if you make sure to have your settings set to not import duplicates. Network discovery can automatically ignore duplicates.

Add these settings are in the network discovery setup correct? Or is this a general setting?

These settings should be in the wizard. When you use the wizard to setup a discovery you should see it as part of the setup. If they happen not to be there they will be on the screen after the discovery is finished. The screen before importing allows you to check if you want to import duplicates or not. I know that by default duplicates are turned off so even if you click import without checking you should see the wizard indicate that duplicates where not added.

One more question. Since I have 2 polling engines, if i run a network discovery via a subnet and have it import on 1 engine, but the device is already monitored via the second, will the discovery import say a new volume that was added?

Example: Below is device being monitored by our primary polling engine.

I ran a discovery, for the subnet the above device lies on, to be imported to our secondary polling engine.

Now the result of the discovery was it skipped this device since it was already being monitored. However, will the discovery be able to import say a new volume that may have been added previous to the device, or will it skip it entirely because the device is already monitored on the primary engine?

Let me bring on adatole​ to help answer this specific question or aLTeReGo​. I'm not exactly sure 100% and wanted to make sure your given a correct answer on this. I myself had similar experiences and that's why I can't answer this specific question myself.

Hopefully we'll have an answer from two amazing experts, gurus in their fields and highly respected individuals in the solarwinds community!

Short answer:

Scanning a network is poller-agnostic - the scan just happens (usually from the primary poller). Once nodes are discovered you can assign those nodes to your various polling engines.

If you scan a node which has already been discovered, it will appear in the scan as "duplicate nodes" and the default is to ignore them.

However, you CAN choose to still onboard the node (with a different name) and assign it to another polling engine if you want some kind of triangulation or double-check for the device.

HTH.

- Leon

Thanks for the reply. My main objective is to have the Network Discovery find and manage any new volumes, of already managed devices, automatically; So that I don't have to add manually. When you set up a discover you have to choose which Polling Engine you want the device to be added to, once its been discovered.

So a node will be skipped, regardless if it finds new volumes, if its already being monitored by a secondary polling engine and the discover took place on the primary (or vice versa)?

I can confirm that if you have multiple polling engines with overlapping discovery profiles created on each, that you will not end up with duplicate nodes monitored by each of your polling engines.

That's good to know. However, my main concern is that new volumes of already managed nodes are skipped if its already be polled on a different engine.

In that instance the Node already exists and is assigned to a specific polling engine (determined either on the original import, or manually changed in the UI). So regardless of which polling engine you select from the discovery wizard, when existing nodes detect changes (such as new volume) it would import the volumes to the existing node, and it will NOT move the Node to a different polling engine. Any devices that are new Nodes will end up on the selected Polling Engine.