Jump to content
Klaymore

HP Proliant Monitoring (SNMP and Insight on Windows )

Recommended Posts

Standard Disclaimer Applies

This is an attempt to document how I am monitoring SNMP for Proliant servers through Labtech. It is a merger of the previous 2 threads on the subject along with my notes from setting it up and muddling through SNMP after years of not looking at it at all. Forgive anything I put in wrong, and please comment it so I can change the content and wording to be as exact as possible.

 

NOTE: This is not my whole solution, but should be enough for anyone to build out their own. Once I figured it out I created all the monitors I wanted ;).

 

 

Process:

  1. Make sure the SNMP feature is installed
  2. Configure the SNMP service
    • Make sure all the boxes are checked and set the contact and location information -- THIS STEP MUST BE COMPLETED MANUALLY ON EACH MACHINE
    • Set up your community strings (one has to be read/write for insight, mine autocreated; one read only, for testing purposes I just used public,admin,proliant for my test values, change for production)
    • Set it up to accept from the local host
       
  3. Install The Insight Management Agents and WBEM providers from HP and their prerequisites
    1. ilo 3/4 Channel driver for server 2016 webpage  (prereq)

    2. iLO 3/4 Management Controller Driver Package for Windows Server 2016 webpage (prereq)

    3. HP Insight management Agent x64 download webpage

    4. HPE Insight Management WBEM Providers for Windows Server x64 webpage

    • My preferred method for this is to setup the standard build for the server (OS and features wise) with SNMP enabled and configured and then run the SPP on the machine and upgrade everything, etc.

  4. Create a group to add your remote monitors to.

    • Add a known proliant server to the group for testing, preferably one you have quick physical access to.

  5. Start the SNMP service, and all the insight management agents listed as dependencies on the SNMP service. 

  6. Add the MIB files from the HP SIM MIB kit. 

    • I found this website to be more helpful than manually trying to read through those MIB files (The link is to the MIB tree that contains the drive array status). Find 'condition'. For now I am only using Scalar Integers with fast and easy go/no-go criteria.

      • Update: My Manual Run script pulls the end OIDS from a table and creates a monitor for each of them, food for thought.

    • The Only ones I actually had to add for this were CPQHOST-MIB, CPQIDA-MIB, CPQHLTH-MIB if I remember correctly. But why not have those OIDS for when your probe walks and to use for detection templates!!

    • See this thread on how to fix up those MIBS so they don't get pesky on import.
      Labtech's guidance on loading up those pesky MIB files.

  7. Add remote monitors against your group(s) on what you'd like to monitor. I added one for the SNMP Service against my mid level SNMP/Windows group. In addition to the following on the HP Proliant sub-group:

    • I did the alert continuous alert style and made the subject something like "%name% %state% on %computername%" or "Proliant SNMP Monitoring Dependency failed" for success and failure. I then input in the failure/success messages what the return conditions mean to aid in troubleshooting.

    • Remote monitors for these services at a minimum: CpMgHost, CpqNicMgmt,CqMgServ, CqMgStor (Name Convention: SVC - %servicename% (SNMP))

    • Remote Monitors for each OID you want to monitor (Name Convention: SNMP - HP Proliant - Drive Array)

  8. Test your monitors (pull a drive to degrade the array! unplug a power supply! hit a drive with a hammer before installing it! perform other testing actions!)

  9. One you have verified this is working on your test server, then I would import the XML scripts in the following order:

    • EDFS in any order

    • Scripts - I set this up so you have 3 groups (SNMP>Windows>HP) on the windows group you schedule run the scripts that install SNMP and configures it daily, and autojoin using search for EDF 'Use SNMP' and server OS, windows (not in my included scripts). Then on the HP Group the magic happens; there is the master script that first checks to make sure SNMP is marked as installed, then enabled, then installs all the software in order using the latest download links I could find. If it is already configured the script exits with no ticket. If windows can't get SNMP, you get the 'general failure' ticket. If any software fails, you get 2 tickets; one for the offending software, and the general one. So with only 3 scripts scheduled, it installs SNMP, configures it, and then installs all the prerequisites for SNMP monitoring on these devices.

      • You have to setup your own ticket comments and finishes. For the application install scripts make sure they are all 'Isolated' and are using the latest EXEs (they get updated about every quarter or two). 

      • I used the following community strings on my test machine, search for references to them and change them to your production key like I did!: public,admin,insight,proliant

      • You have to go into the Manual Run-% script to setup your community key in the SQLINSERT statement. This script creates remote monitors for each physical drive as reported by SNMP.

    • Remote Monitors -- I will export some of the remote monitors I make. They are setup currently to monitor:

      • Drive Arrays, Temperature System, CPU Fans, ASR, Resilient Memory, System Fans, and a few others. 

      • Recommend importing against a known proliant machine computerid (yes you have to modify each file, I set the computerid in the export to %computerid% on all files by opening them all in notepadplusplus then doing a replaced against my computeridfield )then drag to your groups.

  10. Add the searchs to the groups to add the rest of your servers.

UPDATES:

Added in hyperlink to Darren's MIB import fix batch file

Changed wording to be more exact

Added in what services I find it necessary to monitor

added links

Added a verbose roadmap

added some scripts, edfs, searches, etc that I am using.

 

Credits:

Myself

@DarrenWhite99 : Link inside on MIBs and help making the other things

@Joe.McCall : https://www.labtechgeek.com/topic/2827-hp-server-hardware-monitoring/

 

@HickBoy  : https://www.labtechgeek.com/topic/3756-hp-smartarray-monitoring/?do=findComment&comment=22997

 

SNMP HP Proliant.zip

Remote Monitors.zip

Physical Drive Ticketing (export).xml

Edited by Klaymore
Updated
  • Like 1

Share this post


Link to post
Share on other sites
Hey guys,
 
I wanted to piggyback on this post because it is so recent. We monitor our servers using SNMP but this appears to have been removed from HP Gen 10 servers. Is anyone actively monitoring Gen 10 servers? How are you doing it? We used to use the System Management Home Page and Remote Monitors on specific OIDs but that feature isn't available anymore either. Specifically looking for Power Supply Redundancy and RAID alerts/errors.
 
Thanks in advance,
Brett
Edited by bwoodley

Share this post


Link to post
Share on other sites

Brett, this should still be active in gen10 servers; I know for sure you can go the 'agentless' management route through the iLo side for SNMP/WBEM, and am pretty sure they still have the insight manger and insight management agents too. Though I could be wrong on 2/3. I dont have a gen10 server so I've never looked into them much.

 

Share this post


Link to post
Share on other sites

Hello guys - 

Let's start this off with, I am stupid, or at least feel that way when it comes to doing anything outside of the default setup of Labtech. 
I'm trying to utilize some of the SNMP monitoring such as the above, but am clueless when it comes to importing and getting things to work.  The doc makes this seem simple, but I've tested importing one of the remote monitors, but cannot for the life of me, find it anywhere in LT.  I imported SNMP - HP Proliant - CPU Fans.sql as an example, but have not yet been able to locate this inside Labtech.  
Anyone that can point me in a direction?  I would be immensely appreciative.  

Share this post


Link to post
Share on other sites

@myohnk Sorry I dont log in often, but am usually in the slack.

Those SQL files are generalized versions of what I built out. Not for direct importing unfortunately. You have to edit the SQL to update the %computerid% to be a known computerid.

Share this post


Link to post
Share on other sites

I didn't see any other mentions since last year about Gen10, but yes, this is still applicable. The only difference on Gen10 is that it is no longer using the Insight Management Agent. The System Management Assistant Service must be installed and then enabled (it is disabled by default).

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×