This discussion has been locked. The information referenced herein may be inaccurate due to age, software updates, or external references.
You can no longer post new replies to this discussion. If you have a similar question you can start a new discussion in this forum.

Installation Help Needed

Hi Masters/Techies/Readers.

We have installed following modules

NPM for ~1300 Elements,/~4000(~1300up,       ~2700down) interfaces on Machine-A(Vm of Hyper-v)

SRM for ~200 elements on Machine-A(vm of Hyper-v)

Virtual Mangr ~2 Hosts on Machine-A (Vm of Hyper-v)

Infrastructure is about 150 switches,  ~1300 end points,  ~4000-6000 interfaces (up, down,  Active,  unplug),  2 hypervisors,  20 servers,  2-5 dhcp,  1 dns...etc

Machine-A:

Guest installed in Hyper-v.

Windows server2012 R2

Intel xenon 2.4 cpu

Ram 32gb,

Ip:10.10.x.a

Machine B:

Guest installed in Hyper-v

Window server 2012 R2

Intel xenon 2.4 cpu

Ram 32 gb

Apps installed: SQL standard for primary polling engine

Ip:10.10.x.b

Now we want to install following additional modules.

NCM for ~200 devices

NTA for ~1300 devices,  2 applications

IPAM for ~1500 IPs

Please give me an advice for a healthy installation. Some sources said that NTA is heavy loader,  some said installing in vm is not a good choice. Iam confuse in this all.

  • I'd assume that the concern for NTA install as a VM was based on the bandwidth requirements, however NTA isn't super bandwidth heavy. You will need a third server, though, for NTA's dedicated flow storage database. We've run our 3 servers (with a similar scope) as a VM with no problems. In fact, we use a shared SQL database, as opposed to a dedicated one, so you've got us there. emoticons_wink.png

    If you look at some of the installation documents ( NTA Installation - SolarWinds Worldwide, LLC. Help and Support for example) you'll see their recommendations for a separate flow storage server.

    If you are concerned about bandwidth/resource utilization, I would suggest slowly ramping into your NTA deployment - don't add every node all at once, and if it starts eating up more resources than you expected, you can always investigate a physical server alternative while maintaining a working VM.

    HTH!

    EDIT: I'd like to also state that I have nearly all of the same modules - NPM, NCM, IPAM, NTA, and VNQM - with slightly less resources per machine than you (and about the same amount of nodes/elements) and working fine with no bandwidth or resource concerns.