designerfx · observability architect · ✭✭✭✭✭

A good thing frak​, you may want to edit 1. Browse to C:\ProgramData\SolarWinds\Collector\Data\ to say Browse to %programdata%\Solarwinds\Collector\Data for folks who install on other drives such as from following KMSigma​'s scripts or even people who just (thankfully) don't install everything on the C drive.

michal.hrncirik Lync support for call data, etc is the only defining factor aside from Avaya support that will decide whether or not I am able to get this product purchased, so if that can be pulled in would be useful, to say the least. SAM doesn't really cut it here, and the only other option appears to be queries against…

Please let me know. Web isn't my specialty and I don't really have anyone here to assist on that one

trying to revisit this today. I wonder if cahunt​ has ever ended up looking into this further?

I believe I had a discussion with rschroeder​ about #3, where some of the modules required separate VM's/servers for APE's - notably UDT if I recall correctly. I don't think it's just NPM/SAM anymore though, if i I recall correctly.

Can't handle very many nodes? You do realize that if you are polling 12-14k nodes, this calls into question the exact definition of "Very many"? 12-14k nodes (it's not limited to 12k as noted in this very thread) is indeed "very many". Not the biggest, but bigger than plenty of solarwinds customers. Who else does stacking?…

Microsoft has a CAL, they don't sell you unlimited at all. So you're paying "per node", basically. Solarwinds isn't charging you per node with SLX. They'll let you put on as much as you can, at what is easily a fraction of what anyone else charges. Stacking of mailbox servers is called clustering, and it's not as simple as…

dfairles​, your argument is associating hardware with licensing. That's not how this works. You don't *have* to buy more APE's, if you keep beefing up the hardware to the limit of the engine's capability - which will realistically not be worth it after some point, depending on your HW budget. This is not at all like saying…

I think the concept that byrona​ is discussing vs what dfairles​ is saying are pretty different. I do agree with moving away from elements, but that's why it's not displayed anywhere aside from as an admin setting (because most users care about nodes and not # of elements vs us as sysadmins/admins). I think what solarwinds…

monitoringlife​ they may have had bugs to fix, or maybe it didn't work right or something. Never know!

This is it, folks. This is a huuuuuge amount of awesome updating and refining and cleanup and new stuff. Keep up the good work solarwinds!

aLTeReGo​ maybe DanielleH​ / wabbott​ can get this fixed?

xpowels remember that the order listed in the roadmap top to bottom is in order of priority. That means they'll try to put as much of those as they can in the next release/release(s). If it's not on that list but they're interested, it's probably going to be the release after/later. Nobody knows timeframes for this stuff…

andreasw​ unfortunately I know this from having experienced it firsthand. Which is weird because the NAM module (yes, that's the one) is more focused towards Cisco Prime, oddly. Although I do understand when you think about the volume of traffic applied to a 7K and potentially the amount of flows generated.

The only way to get NBAR on a nexus, andreasw​ is through buying that physical application module for the 7K that does a bunch of application stuff. I forgot what it's called.

This fails for me. Are you missing the part of the select query to account for NodeID? Because that's what I'm seeing: "cannot resolve NodeID". Or, are you meaning to use RelatedNodeID? I had to delete the two lines at the top and bottom referencing NodeID to make this work.

One can dream, but I'd love to see NTM integrated into NPM and refreshed accordingly.

Unified installer for main poller yes please aLTeReGo​! Also I see: * Integrated search on this list. Is this something along the lines of what global search was doing?

Oh crap. I just realized this. I saw the 100k/500k thing but didn't realize that means 5 AP's that aren't stacked. stevenwhunt​ has anyone looked at why UDT doesn't do stacking per https://support.solarwinds.com/@api/deki/files/9924/ScalabilityEngineGuidelines.pdf ?

You do it next time because I believe either this update or a future update lets you update pollers concurrently per stevenwhunt​ ? I don't remember if that is existing feature or "we're working on it"

This is my post of jealousy I can't even imagine what I could do with 17 polling engines right now. I worked in a very spread out environment before and being able to spread out the polling engines + EOC to accommodate that infrastructure sounds very useful.

It's an intriguing "big customer" discussion, but by big customer I basically mean NPM/SAM/NTA unlimited with maybe 5-10k nodes depending on use can absolutely be enough to need 2 pollers stacked or even 3 in some cases. I could see a customer with a huge amount of elements (well more than 5-10k) getting by with 3-4 triple…

xtraspecialj That's because we oversimplify by saying elements - for anyone who's looked at the polling engines page can attest to and your post basically answers. We know that there are different kinds of elements with different kinds of loads that reflect on different job weights among other things. It's not an exact…

jbiggley​ solarwinds official recommendation for multiple years has been 10k elements per polling engine, but that's why stacking exists - so you can get 20k elements on a single server (and not worry about # of engines).

literally as ecklerwr1​ is mentioning, you run the additional poller installer on the main box. You can do up to 2 additional pollers (3 stacked pollers total) on the same NPM server. It's a matter of how you build your environment as to if you do it, but once you hit about ~10k-12k elements you can stack your poller and…

or maybe aLTeReGo​ can update it?

Well I know we have EPL but I don't know enough about networking beyond suspecting that it's not sufficient to handle this. I'm new to the FOE side - in fact, I can't even implement it until we get the hardware upgrades we're waiting on.

That's a bit of an issue for us. It's pretty much a requirement for us that we have a WAN DR solution and if they're going to deprecate it we're screwed.

I still don't fully understand why we have: HA, FOE, Active/Active? being 3 separate licenses. That's pretty confusing. Especially considering that FOE says it can provide HA as well. Is that not how this works?

Better than how I had to do it for azure, where I had to build a proxy into a jumpbox for each cloud!