NNMi/NA integration - new discovered node not synced/created in NA

Hello,

NNMi/NA 2022.05.

I came across a strange issue. I was reported that a few new nodes discovered in NNMi where not created in NA. At the same time (the same day) another node was created successfully in NA.
Now I am trying to find out the reason why some of the nodes were not synced to NA (not created).

These are the only entries found in NNMi logs (nnm-trace log is configured with the FINEST level of logging).

/var/opt/OV/log/nnm/nnm-trace.log:2024-08-14 16:36:39.155 INFO  [com.hp.ov.nms.disco.config.DiscoveryConfig] (Thread-83 (group:HornetQ-client-global-threads-474315718)) Seed dbrn41007 status is set to In progress.

/var/opt/OV/log/nnm/nnm-trace.log:2024-08-14 16:40:56.395 INFO  [com.hp.ov.nms.disco.config.DiscoveryConfig] (pool-1-thread-9) Seed dbrn41007 status is set to Node created.

/var/opt/OV/log/nnm/nnm-trace.log:2024-08-14 16:40:58.792 INFO  [com.hp.ov.nms.disco.analyzer.xdp.XdpAnalyzer] (pool-1-thread-9) Delegate to ExDPAnalyzer for: dbrn41007

Is there another place (logs) I can take a look for hints in NNMi or NA? The NNMi/NA integration config is fine - as I mentioned some other node was created in NA after being discovered freshly in NNMi.

Is there a was to force re-sync/re-creation of a node from NNMi to NA?


Parents
  • Verified Answer

    +1  

    Hi, let me see if I can help a bit with this issue.

    First question, if you look at the nodes that weren't synced to NA, are they in the node group for being sent to NA?  

    If they are in the node group, then if you look in NA, can you do a search for the management IPs of those nodes?  I have had times where one way or the other, the devices are in NA (or those IPs are in NA) and so while NNMi is trying to send them over, NA gets the message to add them, but says "sorry, I can't add them, those IPs already exist).  

    Were there a number of nodes added to NNMi (perhaps similar or perhaps different device profiles) and some were added and others were not.  I'm asking to see if NNMi was able to send some of the newly added nodes, just not all or you added only these nodes and they haven't been synced yet?

    Have you noticed this behavior happening frequently, or this is the first time you've seen it?  

    If you look at the NNMi - NA Integration Configuration form, what is your Topology Synchronization Interval?  And how many nodes do you have in total / in the node group for NA?  How long ago were these nodes added to NNMi?  

    I've seen of an issue where NNMi doesn't send nodes over when you expect, for us, we had to increase the topology sync interval, and this helped. 

    You can save the form, that should kick off a new cycle to push to NA.  I'll look for notes on NNMi logging, but I can say you can turn up logging on the NA side and you should see the activity from NNMi.  

    NA logging:

    api/cli/nnmi

    feature/nnmiintegration

    set these to trace and then tail -F the NA log (depending on version - appserver_wrapper.log or jboss_wrapper.log) - you should see what NA core is getting the NNMi updates again frorm the NNMi - NA integration form.  Normally, I'll do the tail -F to a separate file and then in another window I'll watch it.  

    Examples of what to look for in the NA log:

    input: show device -ip "10.85.150.230"

    input: add device -origin "NNM" -ip "10.85.150.1"

    Device could not be added

    show nnmi node association -deviceid "7861" -nnmisystemid "28324c37-478b-4883-b284-5f6d86da9647"

    input: mod device -newip "127.20.210.48" -deviceid "220144"

    input: del nnmi node association -nnminodeuuid "789e157c-72ad-4889-a70c-169dd0968d80" -deviceid "1621" -nnmisystemid "28544437-378b-4883-b284-5f6d86da9647"

    show server option -name "3rd party/integrations/nnmHost"

    For NNMi, this may be out of date, but here's what I had:

    Increase the NNMi logging levels:

    Log onto the NNMi server.
    Backup and then clear the content of the %NnmDataDir%\log\nnm\nnm-trace.log file
    Run the following commands:
    %nnminstalldir%/support/nnmsetlogginglevel.ovpl com.hp.ov.nms.sdk FINEST
    %nnminstalldir%/support/nnmsetlogginglevel.ovpl com.hp.ov.nnm.na FINEST

    Reproduce the problem.
    Note:
    - In order to force a synchronization update for all the synchronization node group, go to NNMi - Integration Module Configuration - NA and re-submit the same data.
    - In order to force a synchronization update for a single node in NNMi:
    Login with the NNMi admin to https://<nnmi_IP_or_FQDN>/jmx-console
    Click on com.hp.ov.nnm.na.im - mbean=NnmNaModule
    Under the resyncNode operation add the NNMi node name and invoke.

    Collect the logs

    Hopefully this will help with troubleshooting.  If you have any questions, just reply back.  .  

    Good luck!

    Chris

  • 0 in reply to   

    Hello Chris,

    thank you very much for this comprehensive set of steps to investigate the issue. Sorry I didn't reply earlier but was occupied by other tasks and what is more strange these missing nodes were eventually synced to NA. The big sync delay which didn't follow the NNMi/NA sync settings still poses some questions but the issue seems to not exist anymore (but I'm having some other issues with NNMi so life is not boring xD).

    Again. thanks a lot. I will save your handy hints as it might I will need them later.

    Best regards,

    Jarek

  • 0   in reply to 

    If you haven't done so already, you may want to check with support on patching.  Not sure if this lines up with your version of NNMi, but might be a good idea:

    Patch / CR List (microfocus.com)

    I know some others who had changed the integration timing for NA (to make things quicker) ran into a similar problem.  It worked fine for a while, however, there would be long gaps where nothing would come across.  My understanding is this is due to how things work within the integration and there's a bit of a problem if you fall behind in the integration queue.  

    Good luck!

    -Chris

Reply
  • 0   in reply to 

    If you haven't done so already, you may want to check with support on patching.  Not sure if this lines up with your version of NNMi, but might be a good idea:

    Patch / CR List (microfocus.com)

    I know some others who had changed the integration timing for NA (to make things quicker) ran into a similar problem.  It worked fine for a while, however, there would be long gaps where nothing would come across.  My understanding is this is due to how things work within the integration and there's a bit of a problem if you fall behind in the integration queue.  

    Good luck!

    -Chris

Children
No Data