Allow NPM to rediscover a node entirely without re-adding.

I did an initial discovery and forgot some of the interfaces on the nodes. Reopened the initial discovery and added the additional interface information. Re-ran the discovery. No interfaces are being added. Looks like you do in fact need to delete the nodes and rediscover them from scratch. I use a lot of custom properties so it's a lot of work to delete the nodes, rediscover, re-import the custom properties, etc. Wish it would actually rediscover the node.

I've submitted a White Paper for Thwack Administrator to consider; the topic is Solarwinds' recommendation for removing "unknown interfaces" caused by replacing a node with a different model while keeping the same name and IP address.

Yes, the recommended process remains deleting the node and re-adding it, when anyone who sees the "Rediscover Node" button would reasonably expect it to do just that--eliminate interfaces that are missing and add new interfaces.

We'll see what happens.

I brought this issue to tech support once before. Their stock answer is to delete the node and re-add it.

Similarly, I replaced a pair of 7609's with a pair of 7009's that took over all SVI's from the 7609's and also their management addresses.  A new discovery showed the new hardware & interfaces, but left all of the original 7609 interfaces in place but grayed out.

Since the old interfaces' data is useful, I kept them for a year.  But it was not user-friendly to have to pull up a grayed-out interface and compare it with the new one.  A nice solution would be to enable correlation of the old interfaces with the new ones--tie them together with a nice gui that shows element 12345 is now interface 33001, and allow the two lines of stored data to line up with each other seamlessly.

Instead, I eventually had to delete the current 7009 node that had the invalid 7609 interfaces still listed, then re-add it and rediscover it.  In the process I lost all the interface history from the 7009.  That's not ideal, and I'd love for SW to find a process to handle replacements and upgrades without losing data, and without doubling the number of elements by keeping elements from retired devices while adding new elements for the replacements.

2nd time voting on same issue. Is this going to happen or ?

This definately needs to happen. Here is a link to a thread mentioning more issues with things needing to change that didnt.

We have a similar issue with AIX systems and drives.  When they do a "refresh" all the drives go into unknown status and we have to delete them and add the 'new' ones....  then we lose all the history for all the volumes and memory