Kaseya Community

Monitoring HP Server Hardware

  • If you have HP Servers you look after (e.g. ProLiant ML, BL and DL family), and you install the HP WBEM providers, you can monitor for hardware-level failures (Fans, Temperature, Disks/Array, CPU & RAM, Network, iLO, etc) with Kaseya. I find this REALLY handy for detecting bad fans and failing disks in RAID arrays - things that the Windows OS can't detect by itself.

    Just cook up a custom monitor event set as follows:

    <?xml version="1.0" encoding="ISO-8859-1" ?>
    <event_sets>
      <set_elements setName="HP WBEM Providers" eventSetId="47983158" snmpTraps="0">
        <element_data ignore="0" source="HP CPU Threshold" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP Disk Threshold" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP Ethernet" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP Fibre Channel" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP Memory" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP MgmtProc" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP Processor" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP SAS" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP SATA" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP SCSI" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP Sensor" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP Smart Array" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP System" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP Tape" category="*" eventId="-1" username="*" description="*"/>
        <element_data ignore="0" source="HP Test" category="*" eventId="-1" username="*" description="*"/>
      </set_elements>
    </event_sets>

    Apply the monitor set to Monitor the system log (Monitor -> Agent Monitoring -> Event Log Alerts -> log type System, Define events to match or ignore - press Edit then import the script above) for any events of type Informational, Warning or Error, and choose whatever action you desire (I use email to my critical monitoring address), but you can trigger all the usual stuff e.g. alarm, create ticket, send email, etc.

    You will also need to turn on logging for the event logs under Agent -> Event Log settings, and ensure that for the System event log, you have types E, W and I enabled.

    I also have a .VBS that I can run on a schedule or as frequently as desired, that queries the hardware via WMI. If you're interested, please reply here and i'll post some more about it.

    The HP WBEM providers software you will need, is at http://h18004.www1.hp.com/products/servers/management/wbem/providerdownloads.html - or just Google "HP WBEM Providers". If you've been using a pre-installed HP OS and/or the smartstart CD, the software you need may already be installed on your server. Look for "HP Insight Management WBEM Providers" on your start menu.

     



    [edited by: chart71 at 9:45 PM (GMT -7) on 4-23-2012] Corrected details, spelling
  • Hello Chart71!

    Great work! Did you have any further information about monitoring these parameters from within a esxi host?

  • Hello Kai,

    You can use Network Monitor for that.

    Sander

  • Hello Sander.

    You mean KNM? KNM isn't a core feature ... i'll use the core monitoring feature and will not buy an additional product... :/

  • HP wbem.zip

    Hi all.

    No, my monitor set won't work in an ESXi environment - because the HP WBEM providers don't work inside a VM. To monitor an ESXi host, use KNM, which as you discovered, is an additional pay-to-use module. I don't have a work-sround for that.

    I've attached my VBScript and agent procedure that gathers the base data about a server and writes it to the script log. FYI, the possible return result codes are as follows (You could expand the script to automatically detect problems and raise tickets, etc. if you like):

    OperationalStatus contains the computer system consolidated status and indicates the worst status of connected subsystems.

    •  0 (Unknown)
    •  1 (Other)
    •  2 (OK)
    •  3 (Degraded)
    •  6 (Error)

     

    StatusDescriptions contains one of the following corresponding to OperationalStatus:

    •  One or more of the subsystems that make up this computer system have an OperationalStatus value of Unknown.
    •  One or more of the subsystems that make up this computer system have an OperationalStatus value of Other.
    •  OK
    •  One or more of the subsystems that make up this computer system have an OperationalStatus value of Degraded.
    •  One or more of the subsystems that make up this computer system have an OperationalStatus value of Error.

     

    HealthState

    •  0 (Unknown)
    •  5 (OK)
    •  10 (Degraded)
    •  20 (Major Failure)

     

  • I can monitor ESXi servers in hardware using www.claudiokuenzler.com/.../check_esxi_hardware.php

    The script works equally well for IBM, Dell, HP and Intel servers - bonus!

    Basically, on any machine that can 'see' the ESXi host, install Python and the WBEM extensions, then run the python script (pipe the reseult to a text file), then capture and process the text file within Kaseya in the usual fashion.

    Overall this really works well. I leave it up to the reader to best determine how to securely store and pass the ESXi login details, and process the resulting file into something useful in Kaseya.

  • Hello Craig,

    just playing around with your idea. It works great!

    A step further it would be a great feature if we could audit networks and get a information about missing wbem providers on hp servers. do you have any ideas how we can creat such a awesome report?