10 Replies Latest reply: Mar 7, 2012 9:14 AM by netlogix RSS

SAM - Adding hardware monitoring to non-ESX hosts en mass?

JedTheHead

I have seen many discussions on Thwack regarding enabling (even auto-enabling) ESX host hardware monitoring within SAM and was curious if there is a way to auto-enable hardware monitoring for non-ESX physical nodes already in SAM?  We have 900+ HP servers already available and working with HPSIM agents and want to turn them all on inside of SAM.  Any easy way? (I'm more than comfortable flipping a field or two inside the DB if needed as well.)

 

(Sorry in advance if this is a re-post, I didn't find one though...)

 

Thanks!

Jed

 
  • Re: SAM - Adding hardware monitoring to non-ESX hosts en mass?
    DONDERKA

    Hi Jed

    is this fresh installation of SAM 5.0 RC or it is upgrade from APM 4.2?

    If you did upgraded APM 4.2 to SAM 5.0 and you had already added those HP nodes before.

    Run Network Sonar Discovery where you can specify IP Addresses which you want to poll for Hardware info. If those servers respond SAM that they are capable to supply Hardware Information's, SAM will automatically enable polling of Hardware Health Info.

    If you have fresh installation of SAM 5.0 RC X.

    The Hardware should be polled from the moment you add new node which is capable supply hardware information's. 

    • Re: SAM - Adding hardware monitoring to non-ESX hosts en mass?
      JedTheHead

      This is an upgrade to the latest RC from APM 4.2.  I will try the sonar discovery on a small batch and report back.  Thanks!

    • Re: SAM - Adding hardware monitoring to non-ESX hosts en mass?
      JedTheHead

      This appears to have done the trick and those nodes scanned now have hardware monitoring enabled.  I will try a larger group of nodes now.

       

      Thanks!

    • Re: SAM - Adding hardware monitoring to non-ESX hosts en mass?
      JedTheHead

      Well, I'm closer.  If I follow the scan method and verify both my SNMP community strings and the Windows Credentials asked for in the sonar wizard, all the hardware monitors are checked but result in an unknown state with the following: Hardware polling failed: Value cannot be null. Parameter name: credentials

       

      If I take the same nodes and manually enable the hardware monitoring via the "List Resources" tool, everything works as expected and the hardware monitors work. 

      Any ideas of what credentials it is looking for?

       

      Thanks!

      • Re: SAM - Adding hardware monitoring to non-ESX hosts en mass?
        martin.susil

        Could you open a support ticket for this issue and reference this thread. Then please post a support case number back here. We will need to have a look at log files to see what exactly is going wrong on your box.

        Thank you

      • Re: SAM - Adding hardware monitoring to non-ESX hosts en mass?
        martin.susil


        Well, I'm closer.  If I follow the scan method and verify both my SNMP community strings and the Windows Credentials asked for in the sonar wizard, all the hardware monitors are checked but result in an unknown state with the following: Hardware polling failed: Value cannot be null. Parameter name: credentials

        Thanks!

         



        We have confirmed this is a bug and it happens if user runs a Network Sonar Discovery after upgrade with following options:

         

        • IP range of existing Orion nodes
        • Do not specify SNMP community strings or only specify invalid SNMP community strings
        • Specify valid WMI credentials.

        Then APM incorrectly wants to use WMI to poll data from this SNMP node. We are tracking this internally as FB110354. 

        Workaround  is to click on 'List Resources' button on affected node, select 'Hardware Health' and then submit the changes. Unfortunatelly there is no way how to fix this for multiple nodes at once. If this is occurring on many nodes and it would be too difficult to fix it manually for each of them, then please open a support ticket.

        Thank you

        • Re: SAM - Adding hardware monitoring to non-ESX hosts en mass?
          JedTheHead

          We have an existing case open on this #318282.

          Thanks!

          • Re: SAM - Adding hardware monitoring to non-ESX hosts en mass?
            netlogix

            It sounds like a pretty specific bug, what if, for a one time scan, you added all your possible snmp strings and not supply any windows credentials?

            It doesn't fix the bug, but it gets you past the stopping point.  Maybe by disabling the windows (WMI) credentials, you might get 90% of the servers in.  I don't know the exact specifics of your environment, but this might help.

            • Re: SAM - Adding hardware monitoring to non-ESX hosts en mass?
              JedTheHead

              Thanks for the reply!  Actually, all of our servers are already managed by APM/SAM.  We upgraded to SAM and are attempting to enable hardware monitoring for all 980+ nodes en mass without having to do it manually on each Node - which would take quite a lot of time.

              Thanks!

              • Re: SAM - Adding hardware monitoring to non-ESX hosts en mass?
                netlogix

                the way I enabled the hardware monitoring on my servers when I did the upgrade was I opened Manage Nodes and open by category of physical servers, displayed the IP address, Copied and paste special into excel, then copied them into a one time discovery, ran it (without windows credentials), then imported the already existing nodes (but not the volumes or interfaces due to those silly server 2008 extras I would have to clean up), and that did it.  I agree it's not all that pretty, but it worked for me to enable my hardware monitoring on my physical Dell servers.

                Oh, if you have some monitored by wmi not snmp, you could included the polling method in manage nodes and do one discovery for snmp and one for wmi.  It feels weird doing a discovery and import on nodes that already exist in Orion, but that's also how one had to get topology information before too.