I am not certain, at least not for every different circumstance, however, I would think it best to upgrade your NPM core license first, before adding an APE. The NPM core polling engine is capable of 10k+ elements, with moderate polling cycles, before needing to add anything. All of the technical aspects aside, it most likely would not be wise, financially, to add additional NPM core pollers without maxing out the previous/current poller first. I believe the annual maintenance costs will eventually add up to more than it would have cost to have simply upgraded to the proper licensing count to begin with.
Network Performance Monitor (NPM), License with 1st-year Maintenance
NPM SL100 SolarWinds Network Performance Monitor SL100 (up to 100 elements) Change License
NPM SL250 SolarWinds Network Performance Monitor SL250 (up to 250 elements) Change License
NPM SL500 SolarWinds Network Performance Monitor SL500 (up to 500 elements) Change License
NPM SL2000 SolarWinds Network Performance Monitor SL2000 (up to 2000 elements) Change License
NPM SLX SolarWinds Network Performance Monitor SLX (unlimited elements-Standard Polling Throughput) Change License
APE is $17k, and I believe it is only for unlimited licenses (not the actual unlimited, but an alternate, limited version of unlimited... yeah... weird...) , either because you really would/should not require it until you have maxed out NPM SLX elements, or it matches the max of the NPM license you are using.
Though it may not fully, or technically, or accurately, explain it very well, that is how my brain understands it.
I hope it helps, though, there is a good chance I fumbled it up there...
1 of 1 people found this helpful
Agreed. Having everything on your core polling engine and adding APE's when you exceed your polling is the way to go from everything I've heard. EOC is mean to tie together multiple very large Orion instances and provide a single summarized pane. You don't get the full functionality of a single integrated Orion instance. This might give you some insight... (Re: EOC Still seriously lagging behind)
If you're having performance problems, look at the hardware not only on the core polling engine, but your SQL server also. Problems in the performance of your SQL server will make Orion run like molasses...
Sorry, I left out a few details.
We have an SLX license for NPM. The redesign is due to regulatory compliance which requires us to firm up a particular vLAN (close firewall holes). This vLAN happens to be where our primary NPM instance is located. Right now it polls its own and several other vLANs, and also a few remote sites.
I will be putting polling engines (of some sort) in the remote sites and am also moving the original NPM instance to another vLAN. This is where the concern over costs came in....
Although we are not close to capacity (SLX license with only a few thousand elements now monitored), I still need to put some kind of a polling engine in place on the 'secured' vLAN (and remote sites) and was wondering about the cost vs the actual best practice solutions. I am also installing EOC to monitor this/these and several other remote sites that already have their own NPM instances.
So, from what you've stated it seems that it just comes down to an argument of a single point of management vs cost. I think cost will win out on this one.... oh well.
Thanks for the advise and VERY QUICK RESPONSES!
galvin Well, I suppose you could just as easily get the Engineer's Toolset, and just run NetPerfMon app from that. However, you will only have very basic stats and functionality. I actually have this setup in case we have some major problems. I have all of our nodeIP/nodeNames exported out of our primary db, and imported into the NPM app running from the toolset (it has it's own database). Engineer's Toolset is under $1,500, I think, and having already purchased it, I found it could provide us just that little bit of basic monitoring we could use if we ever had to in a pinch...