HP NNMi 10.00 Released

HP NNMi version 10.0 has been released. This is a good release, with many usability enhancements. I’m pleased to see continued development, as the future nirvana of all-powerful software defined networks hasn’t quite arrived yet. For now, we still have to manage our networks the old-fashioned way: SNMP is still alive & kicking.

NNMi – Background

HP NNMi is a spiritual descendant of HP OpenView, one of the first network monitoring tools. Between versions 6 and 7, HP completely re-wrote the NNM code, and now we have NNMi. The core product performs network discovery and fault monitoring. Add-on components (iSPIs) offer performance monitoring, NetFlow analysis, IP SLA monitoring, etc. A sister-product (HP Network Automation) is used for network configuration management. The add-on components were all separately licensed, but HP now bundles products together.

Historically NNMi has focused on underlying network monitoring capabilities, and less on the user interface. This meant that almost anything was technically possible, but the visual experience was underwhelming. The integration between core product and add-on components was limited.

The last major release was 9.20, in June 2012. There have been minor enhancements and fixes since, but the last patch was in September 2013. We’ve been due for an update.

Key Features

There’s three broad strokes – improvements to usability, cleanup of old code, and better component integration. Here’s a quick list of the most interesting changes:

  • Dashboards – key status and performance indicators at a glance
  • Scheduled Outages
  • Improved chassis/stack management
  • Support for VDCs and server LAGs
  • IPv6 on by default, on all platforms
  • iSPI Performance can be distributed across multiple systems
  • Platform changes – dropping Solaris, HP-UX & RHEL 5.x support
  • Removal of legacy integrations
  • Updated browser support
  • Map annotations – add text to maps

Drilling into a few of those:


The most obvious change is adding dashboards. Dashboards pull together many different status and performance indicators for a node or group of nodes. From one page you can now get a good sense of the health of the network – top interfaces, systems with faults, highest CPU, etc. All that information was available separately before, but you couldn’t pull it together in one place.

Here’s some example screenshots (click for larger):

Network Overview Dashboard

Network Overview Dashboard

Component Performance Dashboard

Component Performance Dashboard

I haven’t been able to show all the widgets on each dashboard – they scroll down beyond one screen here.

This version has limited dashboard customisation capabilities, but I expect that will change.

Scheduled Outages

In earlier versions, you can mark a node as “Not Managed” or “Out of Service.” This would stop any alarms for that node – helpful if you’ve got planned work, or a known outage. The problem is that it was easy to mark a node as out of service…and then forget to re-enable monitoring in future. You might never know there’s anything wrong…until something breaks, and people start asking why they didn’t get any alarms.

Now you can configure a scheduled outage, either for planned work, or when there’s a known outage that you expect to be restored shortly. You can also configure retrospective outages, so the downtime won’t count against your SLAs.

Scheduled Outage Configuration

Scheduled Outage Configuration – click for larger

It doesn’t look like you can scheduled recurring outages, but they can be configured via CLI, so you could script it.

Platform Changes

NNMi 9.20 was supported on Solaris, HP-UX, RHEL 5.x & 6.x, SUSE Linux and Windows 2008 R2. Historically the customers that used OpenView/NNM were the customers that ran Solaris/HP-UX Big Iron systems. But those systems have been mostly supplanted by Linux. These days there is little reason not to run on a Linux VM. There are standard migration paths to move from Solaris/HP-UX -> Linux.

Removing these legacy platforms makes little difference to customers, but it will make a huge difference to development, test and support. Consolidating platforms will cut costs, and should speed up future development.


This is a significant release for NNMi. They’ve taken significant steps to improve product usability. They’re also starting to deliver on the promise of “fully integrated network management.” They’re not quite there yet, but they’re getting closer. I’m also pleased with the changes that remove legacy platforms and integrations. Removing old code makes future test & development cycles simpler and faster.

It’s a major release, so I would hold off upgrading existing installations until patch 1. But I do recommend running it in new deployments, and in your test environment. Note that you will need to re-license NNMi if you are upgrading (no charge if you have a support contract).


44 Responses to HP NNMi 10.00 Released

  1. Michael Procopio August 12, 2014 at 5:09 am #

    I’d love to repost this on the HP blogs would that be OK? If so email me michael.procopio [at] hp.com .

  2. Martin Bilgrav August 13, 2014 at 2:31 am #

    So when can we expect Patch-1 ?

    • Lindsay Hill August 13, 2014 at 8:42 am #

      Sadly I have no inside knowledge, but I would guess a few months away – maybe October? That timing should be OK for those looking to upgrade current instances, as they’ll probably want to run v10 in lab/test first.

      Might be worth chatting to a friendly support contact if you have any? Get a feel for any hotfixes/problems they’re seeing come in.

      • Lindsay Hill December 23, 2014 at 12:07 pm #

        10.00 Patch 1 has now been released – see KM01324914. Lots of added device support, and a few bug fixes in there too. Looks like a good time to move to 10.x.

  3. Bjørn Kvile November 18, 2014 at 1:21 am #

    You did a HP IMC vs HP NNMi comparison a while back.
    There are new versions of both products. Does this change your evaluation ?

    • Lindsay Hill November 18, 2014 at 9:47 am #

      Hi Bjørn

      I don’t think it makes any major difference to the product positioning, but I’ll take a look at that old post, and see if it needs any updates.

      • Bjørn Kvile November 18, 2014 at 9:45 pm #

        Thanks. Especially the SDN positioning is interesting.

        • Lindsay Hill November 27, 2014 at 9:39 am #

          I don’t think there has been any significant shift here yet. If you’re using HP’s VAN SDN Controller, you’ll need IMC. I also expect to see IMC integration with NSX, and OpenDaylight as it gains more traction.

          My expectation is that we will see that with IMC sooner than with NNMi.

          If we see widespread market adoption of a few controllers – let’s say OpenDaylight, ACI & NSX – then NNMi will probably have hooks into those controllers. But I would expect that to be 2016 at the earliest. There might be some limited integration before that.

          This is all based on personal opinion, I have no inside info on it.

  4. Feng Meng September 12, 2015 at 2:43 am #

    Thanks Lindsay, very helpful. Are you aware of SNMP changes between older NNM and NNMi? If I feed NNM or NNMi into an event correlation product, are there any gotchas from SNMP events perspective?

    • Lindsay Hill September 12, 2015 at 8:36 am #

      Not that I can think of, off the top of my head. Obviously the process for configuring forwarding will be a bit different, but the data should be very similar. There are some changes around the defaults for the way things are monitored, but that’s something that should be reconciled when setting up NNMi. Shouldn’t change much from the perspective of an event correlation system. Good thing is that the NNMi series has been out for a while now, and is now well settled down. The early limitations have been removed.

  5. Sarita December 15, 2015 at 10:18 pm #

    Hi Lindsay. Does NNMi 10.10 have capability to integrate with BMC Remedy for purpose of ticketing? Is there any SPi for it? Similarly for Service Now? Thanks

    • Lindsay Hill December 16, 2015 at 9:47 am #

      There’s no out of the box integration with Remedy or ServiceNow. You’ll have to write your own integration. NNMi supports a bunch of methods for either forwarding traps/events, or running custom scripts. I’ve done ServiceNow integration in the past with custom scripts.

      • Sarita December 16, 2015 at 6:56 pm #

        Thanks Lindsay, that was helpful. Would you have any sample template which could provide idea as to how it could be done?

        • Lindsay Hill December 16, 2015 at 8:43 pm #

          No sorry – it was a few years ago I last did it, and it was quite customised for that client.

          It should be fairly trivial to do though.

  6. tracy avent, hpe software instructor/consutant December 19, 2015 at 9:53 am #

    i am told by intenal factor contacts that 10.01 (patch 1 for 10.00) was released in late 2014. So far I have had trouble finding docs about it
    on the support website. I hope that is a defect in the re-designed site and search engine. Investigation continues.

  7. Sarita January 6, 2016 at 9:11 pm #

    Hi Lindsay,
    We wish to monitor certain network elements from a vendor who does not have MIBs for it. Is it possible to monitor objects without MIBs loaded in NNMi?
    Can NNMi produce Availability reports (aka CMDB) and have router configuration backups?

    • Lindsay Hill January 7, 2016 at 10:27 am #


      It’s been a while since I’ve done it, but I think you should be fine adding custom monitoring if you know the OID you want to monitor.

      You can produce availability reports using iSPI Performance for Metrics. Note this used to be a separate product, but a couple of years ago they changed the license bundles, so that all paid customers will have it now. See here for info on the bundles “Ultimate” and “Premium” https://anywebtech.wordpress.com/2014/08/12/nnmi-version-10-0/

      NNMi doesn’t do configuration backups on its own. You can use HP Network Automation for that. HP NA can integrate with NNMi. It is separately licensed, although I think they’ll do bundles if you’re buying HP NA + NNMi together. HP NA does a lot more than just backups though – you can use it for all sorts of network config automation. Quite a nice product, if you can afford it.

  8. Sarita January 7, 2016 at 6:36 pm #

    Thanks for a comprehensive response Lindsay. Have a nice day.

    • tracy avent-costanza January 8, 2016 at 3:53 am #

      do we possibly mean 10.10 ?

      • Lindsay Hill January 8, 2016 at 8:29 am #

        10.10 is the most current release. This post was originally written when 10.00 was released.

        • tracy avent costanza January 9, 2016 at 3:24 am #

          odd, since I am just seeing the item appearing as a new posting with current time-stamps. I gather 10.10 was officially announced this week?

          • Lindsay Hill January 9, 2016 at 6:45 am #

            Which time stamps are you looking at? The main post time stamp says “Posted on August 10, 2014”

            There are a few new comments this week.

            10.10 was released in December 2015, not long before my first reply to you in these comments.

  9. Ram January 20, 2016 at 8:15 pm #

    Thanks Lindsay, Wonderful post!

    By any chance do you have list of all NNM/NNMi versions change log of the features or the licensing model?

    • Lindsay Hill January 20, 2016 at 9:39 pm #

      Best place to look is probably the Release Notes, from softwaresupport.hp.com The 10.00 release notes include a section on the licensing model changes.

      • Ram January 20, 2016 at 9:47 pm #

        Thank you, I will search there.

  10. Sarita January 28, 2016 at 8:19 pm #

    Greetings Lindsay! Does the license of NNMi 10.10 Ult. for Windows vary from Unix, or the key can be used for either of them? I know though the NNM installables are different. For two virtual machines, and all SPIs to be installed, what could be the best suggested distribution on both the servers? Is there any sequence to install?
    Can I follow hardware recommendation for ‘Small’ set up for around 700 nodes?

    • Lindsay Hill January 29, 2016 at 10:48 am #

      From memory the license keys are the same. HP doesn’t care much whether you install on Windows or Linux, although Linux is my preference.

      You should be OK with the ‘small’ recommendation for that many nodes if you’re just doing base NNMi, but as you add iSPIs you’ll probably need more resources. Don’t forget to change Java heap sizes if/when you add RAM. You’ll need to go through all of their release notes to check what’s required. Install process will be something like NNMi on one server, iSPI Performance for Metrics on another server, and then add the other SPIs you’re using. They’ll have pieces installed on both systems.

  11. Sarita January 29, 2016 at 5:19 pm #

    Can hosting all of them on just two systems possibly cause any performance hitches?

    • Lindsay Hill January 29, 2016 at 5:22 pm #

      Depends how much resource you give them, and how much load you put on it – e.g. if you send huge amounts of NetFlow/sFlow traffic, you will have problems.

  12. Sarita January 29, 2016 at 5:37 pm #

    Thanks :). Would figure it out.

  13. Sarita February 4, 2016 at 2:48 am #

    It seems the Metrics Performance and QA SPIs are the way to go. I plan for combination of NNMi and QA one server and NPS and Metrics on another. Would like your suggestion on this please. Does NPS get installed as part of Metrics SPI ?

    • Lindsay Hill February 4, 2016 at 9:34 pm #

      Yes, that is the right way to go. Installing iSPI Performance for Metrics includes NPS.

  14. Sarita February 4, 2016 at 3:22 am #

    Alternatively, if due to hardware limitation, iSPI QA wouldn’t be possible, can the latency and packet loss be monitored through any traps or custom polling?

    • Lindsay Hill February 4, 2016 at 9:34 pm #

      You can do some things with monitoring IP SLA probes through custom polling + traps, but it’s a bit of a PITA to set it all up manually. Adding iSPI QA will make this easier.

  15. dev_7 February 16, 2016 at 9:56 pm #

    Looks like I stumbled upon a very informative blog indeed. I am executing similar install and had a question. Is it mandatory to have shared data directory under C:\ drive? Can it be something like D:\xyz_data\ ? Similarly, can installation be carried in D:\NPS_install rather than C:\ which had Windows?

    • Lindsay Hill February 17, 2016 at 7:34 am #

      Yes, you can have non-standard paths. You can separate app + data directories too if needed.

  16. Sarita May 18, 2016 at 6:09 pm #

    Hi Lindsay hope you doing good. I want to decide if iSPI NET should go into my setup. Can it run with plain NNMi, or I’ve got to have HPOO for it? Does it really provide advantage?

    • Lindsay Hill May 18, 2016 at 6:25 pm #

      It can run with plain NNMi. From memory, it installs a cut-down version of HPOO.

      I wouldn’t bother with it myself though. What problems are you hoping to solve with it?

  17. Sarita May 18, 2016 at 8:09 pm #

    That’s precisely I thought I wanted to know. As I can see in 10.10, it has some features like trap analytics, etc. Does NET provide any further diagnosis which is worth use?

    • tracy avent-costanza June 10, 2016 at 12:44 pm #

      the net spi also provides a “network diagnosis” feature set but that is essentially an integration point with the operations orchestration (OO) product which is separately installed and licensed and typically configured on a dedicated server separate from the NNM platform. I did notice something in the 10.10 release notes that suggested the net ispi might also require a specific license of some kind in order to use OO, whereas the trap analytics is apparently part of the net ispi option license by itself.

  18. MGS August 14, 2016 at 5:52 am #

    Do you know if there is any integration between HP and CitraTest APM – http://www.tevron.com for APM application performance monitoring?

    • Lindsay Hill August 14, 2016 at 7:07 am #

      Nothing for NNMi. Possibly with other elements of the HP suite.