Optimizing Network Sonar Discovery

mstyler

Hello everyone. New to Solarwinds but learning a lot already.

My company is trying to optimize how to run a new network discovery on our Class B Network. With the way it's working now, it would take upwards of 5 days to over a week to discover everything we need, and the list would be incredibly exhaustive. We're trying to manage it by gathering info on 5 class C networks at a time, but the process is tedious setting already discovered nodes to the "ignore" list while trying to gather new information.

My question is this: Is there a way to optimize Network Sonar Discovery to only identify and add to the list NEW nodes, and/or create a means of ignoring everything it has ALREADY discovered. Hope I'm making sense here.

Thanks.

grantallenby

If I'm honest this is something that I have been wondering before.

What I found is that it will discard nodes already added but only after it has discovered them.

Personally I use my old enemy Excel to assist. Im guessing you would have an inventory somewhere? so I would put just the IP's into a column and then run a report in solarwinds that shows all nodes including the IP field and then get excel to highlight any duplicates. That way you are only left with the IP's that haven't been discovered and just run the remaining through the specific IP section in network discovery.

For me I really only use a list of all the IP's and combine it with the section method "specific nodes". But I understand this might not always be an option.

Good pre planning can also help.

Such as removing all community strings that are not needed for this particular scan. If all are needed then separating them up helps a lot. As you can imagine it will check until it finds a working one. This take a lot of time if you have the retry set to more then a few tries.

Also separting devices like servers/ VM's from network devices that are going to have config manager on them. (This one isn't so important and shouldn't have too much of an affect however good house keeping is never a bad thing)

The biggest killer is the time outs and retries in the discovery settings. I personally set those as low as I realistically can.

But warned that I have found it wont always discover everything first try, so don't take it for granted that it will have correctly got everything. I find I'm missing between 5% and 10% every time I do larger discovers say 1000 nodes plus.

Hope it helps in some way.

Grant