DX NetOps

  • Private Community
 View Only

  • 1.  SNMP agent Stopped alarms not auto clearing for 1 particular landscape

    Posted May 21, 2018 02:21 AM

    Hi Team,

     

    We have a distributed setup in our environment , wherein we have 5 primary spectroserver for which we have 5 HA and then 5 DR set for same. For 1 primary landscape , auto clearing is not happening for SNMP agent stopped alarms , not sure whether same condition for other alarms also. Could you please help in fixing this ?

     

    Regards,

    Vaibhav 



  • 2.  Re: SNMP agent Stopped alarms not auto clearing for 1 particular landscape

    Posted May 21, 2018 02:35 AM

    Is this a issue with only Management Lost alarms or all the alarms? What happens if you try to manually poll the device from OneClick?

     

    Please check if your issue is related to the below 

     

    Management Agent Lost alarms are showing in CA Spe - CA Knowledge 



  • 3.  Re: SNMP agent Stopped alarms not auto clearing for 1 particular landscape

    Posted May 21, 2018 03:13 AM

    Hi Phani,

     

    If I am manually pollng these devices from oneclick poll response is successfull.. 

     

    Regards,

    Vaibhav



  • 4.  Re: SNMP agent Stopped alarms not auto clearing for 1 particular landscape

    Posted May 21, 2018 07:13 PM

    What version of Spectrum you are on? Is it only few specific alarms which are not clearing or all of the alarms ?

     

    Did you check if the above tech doc would be of any use?



  • 5.  Re: SNMP agent Stopped alarms not auto clearing for 1 particular landscape

    Posted May 22, 2018 01:28 AM

    HI Phani,

     

    Spectrum Version -10.2.1 ..  Few specific alarms are not auto clearing ..

    Yes i referred tech doc but it did not fixed my issue.. 



  • 6.  Re: SNMP agent Stopped alarms not auto clearing for 1 particular landscape

    Broadcom Employee
    Posted May 23, 2018 09:26 AM

    Hi - so this is R10.2.1 - and a DSS install covering multiple Landscapes - and this one SpectroSERVER is in trouble. I expect you can try to model one of the problematic devices (or are all device problematic?) - onto an other SpectroSERVER/landscape to verify the SNMP agent itself is totally fine (which the poll already points to).

     

    So - next then is - once the SpectroSERVER is restarted - if that symptom persists?

     

    Next then is - to check into $SPECROOT/Install-Tools/.history file - if there are patches already installed.

    Finally - is that device (or these devices) are monitored by SNMPv3 setup?

     

    It appears there are some massive SNMPv3 issues for which corrections already exists - but in case this is not SNMPv3 then there are some fixes around SNMP-polling threading issues.  cheers, Joerg