Systems Management

Forums

Systems Management

Systems Management
Dell Systems Management Solutions: Dell OpenManage, iDRAC, Repository Manager, Microsoft SCCM, Chassis Managment Controller, and more

ESXi 4.1 and OME not getting SNMP, WSMAN is working

  • Hi,

    I have the latest vib installed on my ESXi host

    ~ # esxupdate query
    ---------Bulletin ID--------- -----Installed----- -------------Summary-------------
    ESXi410-Update01 2011-05-02T03:20:23 VMware ESXi 4.1 Complete Update 1
    ESXi410-Update02 2011-11-10T08:09:05 VMware ESXi 4.1 Complete Update 2
    ESXi410-201201402-BG 2012-02-06T05:41:34 Updates VMware Tools
    ESXi410-201201401-SG 2012-02-06T05:41:34 Updates Firmware
    Dell_OpenManage_ESXi410_OM650 2012-02-17T06:02:15 OpenManage 6.5 for ESXi410
    
    
    ~ # /usr/lib/ext/dell/srvadmin/bin/dataeng status
    dsm_sa_datamgrd (pid 6126 ) is running
    dsm_sa_eventmgrd (pid 6666 ) is running
    dsm_sa_snmpd (pid 6680 ) is running
    
    
    C:\Program Files (x86)\VMware\VMware vSphere CLI>vicfg-snmp.pl --server ServerIP --show
    Enter username: root
    Enter password:
    Current SNMP agent settings:
    Enabled : 1
    UDP port : 162

    Communities :
    xxxx

    Notification targets :
    10.138.4.67@162/xxxx
    
    
    
    
    but on my OME server
    When I run the SNMP test in Troubleshooting tool, i get:
    The SNMP read request has failed for Get Community name - xxxx
    Under the machine details - Software Agent Information
    Server Administrator is flagged as yellow with Agent Version showing as N/A
    
    
    Any ideas?
    Regards,
    Lester
  • Hi Lester, thanks for the post.

    First I will point you at this thread for setting up ESXi and OME.  Be sure to have a look.

    en.community.dell.com/.../19424735.aspx

    en.community.dell.com/.../19436722.aspx

    Also,

    for ESXi you just need WSMan, not SNMP for discovery.

    The SNMP part for ESXi only comes into play for "traps", not discovery.  So the TS tool SNMP test is not relevant here.

    Thanks,

    Rob

  • Hi Rob,

    I have read the first link before so like Mikey in the 2nd thread, the SNMP configuration threw me a curve ball.

    I managed to access the OMSA via another server so that's working fine but it's not showing up in OME, so it's being flagged with a warning.

    Regards,

    Lester

  • Lester - I am still in a bit of a pickle with mine, but I might be able to assist you a bit, eventhough I am using 5 and not 4.1

    During your discovery/Inventory - what are you using?  You should only be using WS-MAN  I used Root/Password, checked "secure mode" and "trusted site"

    Give that a whirl....

  • Hey Mikey,

    Yeah I did that already after Rob pointed out that SNMP is only used for traps. WS-MAN is setup like yours and working for most part except that I'm not getting anything for OMSA

    http://prntscr.com/6cs4n

    Regards,

    Lester

  • I misunderstood your statement of "but it's not showing up in OME, so it's being flagged with a warning." meaning that it was still showing as Unknown.

    I on the other hand still cannot access my ESXi5 through another server....

    Being a guinea pig is not fun :)

  • did you try running this from SSH or local tech support mode to check if OMSA is running?

    /usr/lib/ext/dell/srvadmin/bin/dataeng status

  • This looks working as expected to me. We don’t get inventory from OMSA in ESXi so I guess we are not getting the OMSA version and are hardcoding it to “N/A”. The health is showing up as warning which is correct and in fact we show that agent row for displaying the health.

    We are adding more inventory for this in an upcoming release (not the Q1 release).

    Rob

  • Hi Rob,

    Since we are unable to get inventory from OMSA, which includes hardware status. Does it mean that if we have a failure on a harddisk or RAM, there will be no alerts being issued?

    Regards,

    Lester

  • Hi Lester,

    We *do* monitor health which includes hardware health.  We *don't* get other inventory from OMSA.  That is the difference.

    But that will not prevent us from telling you when there is a hardware failure.  The health status shown in the OMSA Agent row is indicative of hardware health.  You can get additional info on the warning status by opening OMSA and drilling in.

    This can be a bit confusing and we are doing more work on this in a subsequent release.

    Thanks,

    Rob

  • Have you had any complaints about high bandwidth usage when using the WS-MAN discovery? I was just able to configure the WS-MAN discovery to get my ESXi hosts yesterday, and today my Network Admin is asking why the management server is using 100 MB/min.

    Any thoughts on that? I realized I had 1 minute status polling turned on (that is default out of the box by the way), so I dropped that down to once a day.

    Do we need to use status polling? Or can we assume that SNMP will continue to alert us when ESXi servers go down or other issues?

    Thanks

  • Thanks for your post.

    Status polling once a minute does not seem normal. The default status polling interval is once per hour. Not sure how it was set to once a minute on your config. We have not heard any WS-MAN related high bandwidth usage issues. If the polling was set to one minute, it definitely explains that usage. I would advise it to set it to once per hour and see if the bandwidth utilization is acceptable.

    Status polling is used for getting the server health information. It allows OME to monitor if the server reports any issues with hardware health.

    Regards

    Abhijit

  • While troubleshooting this high connection issue I have also discovered you cannot highlight a group of discovery ranges and disable them. As soon as you change the page and come back, they are enabled again. You have to highlight each, one at a time, and disable them.

  • Thanks for the update. Something is not right. With status polling turned off, the only other operations would be discovery/inventory which by default once a day usually during night time. Are you running any tasks/updates for any of the targets. I'm assuming you are using OME v 1.0.1.

  • Hi,

    ...really curious...

    When you say you turned the server off over the weekend...which server?  Target or OME server?

    It's funny because the amount of data that we collect from the target via WSMan is not too much at all.

    Can you check the settings again for disc/inv frequency?  It should be set for something like once per day.

    Also, on the portal page, there is a web part for task execution history.  You should be able to view there whether you are having a bunch of tasks run with an unexpected frequency?

    Otherwise, it looks like we will have to ask you to open a ticket at 800-945-3355 so we can collect more info.

    Thanks!

    Rob