Help Please!! Cluster MOnitoring

Hello All,

 

I have OM 8.16 on my environment and I want to monitor vertias cluster application on 2 physical nodes. 

 

I need to monitor only processes on these nodes. So, can you please help me in assisting with this.

 

I have created apminfo.xml

 

<?xml version="1.0" ?>
<APMClusterConfiguration>
<Application>
<Name>CME</Name>
<Instance>
<Name>CME</Name>
<Package> itgsydc11_cme_uat</Package>
</Instance>
<Instance>
<Name>CMS</Name>
<Package> itgsydc11_cms_uat</Package>
</Instance>
</Application>
</APMClusterConfiguration>

 

Now, if it is just the process monitors, how should I proceed?

 

This is the output of ovclusterinfo -a

 

/opt/OV/bin/ovclusterinfo -a

#Cluster

type         VERITAS Cluster Server (VCS)

name         itgsydc11

status       Up

nodes        itgsydc11n1 itgsydc11n2

groups       ClusterService itgsydc11_cme_uat itgsydc11_cms_uat itgsydc11_net_par

#Node itgsydc11n1

state        Online

id           0

#Node itgsydc11n2

state        Online

id           1

#Group ClusterService

state        Online

nodes        itgsydc11n1 itgsydc11n2

local state  Online

active node  itgsydc11n2

#Group itgsydc11_cme_uat

state        Online

nodes        itgsydc11n1 itgsydc11n2

local state  Online

 

active node  itgsydc11n2

#Group itgsydc11_cms_uat

state        Online

nodes        itgsydc11n1 itgsydc11n2

local state  Online

 

active node  itgsydc11n2

 

 

Am I writing the information in apminfo.xml correct? Also if it is only processeses, I am havign a single process montioring for all the processes, how to create app.apminfo.xml

 

Thanks and Reg,
Sree

 

 

 

 

  • If you dont have much idea on apminfo.xml, then I would suggest you to go with the easy way, i.e., creating virtual node.

     

    Add a virtual node and map the 2 physical nodes to it and then deploy the process monitoring policy to the virtual node, the policy gets enabled or disabled based on the state of Resource Group.

     

    For more details refer the Virtual Node Concept in the Online Help.

     

    Hope this helps you.

     

     

    *Edited to correct the Type Error.

  • You have got one more easy method:

     

    Goto the Policy Grouped by Type -> Agent Policies -> Config File.

     

    Open the ClAwConfig Policy and then modify it and then deploy on the physical nodes, it automatically creates the necessary apm files on those physical servers

  • Hi Vidyasagar,

     

    Thanks for the reply.MY OM version is 8.16 . 

    Between I tried copying the apm info.xml and the app.apm.xml to the locations in Linux server. But when I run  ovappinstance, I am getting the below err.

     

    /opt/OV/bin/ovappinstance -vc
    ERROR:    (conf-600) Unspecified exception.
              (conf-666) Application configuration file not found.

    [root@itgsydc11n2 ~]# /opt/OV/bin/ovappinstance -vc
    ERROR:    (conf-600) Unspecified exception.
              (conf-666) Application configuration file not found.

     

    Thanks,

    Sree

  • If possible can you please attach those 2 files.

    I'm assuming that you have placed:

    apminfo.xml : Under /var/opt/OV/conf/conf/

    app.apm.xml: Under /var/opt/OV/bin/instrumentation/conf
  • Thanks a lot,

     

    I am attaching the files..

     

    Also, In my environment. I have 2 physical servers and  2 Virtual Ip's with 2 resource group,. So, How I should be  configuring the nodes in node group. We need only process monitoring on these nodes. Also,  /var/opt/OV/conf/ conf, here the second conf folder was not available and I manually creaTED IT.

     

    Thanks,

    Sree

    New folder.zip
  • Hi,

     

    Try the attached files. 

     

    Keep, apminfo.xml under Under /var/opt/OV/conf/conf/

     

    Control-K.apm.xml and Control-M.apm.xml under the /var/opt/OV/bin/instrumentation/conf

     

    Also try giving the full permission to both files and the folder which you have created.

    APM.zip
  • Thanks a lot Vidya sagar, really appreciate that. 

     

    The user group ahs full access and now the 

    /opt/OV/bin/ovappinstance -vc command is not returning any error.

     

    Other details

     

    1) added the virtual nodes - mappped it to  physical nodes and  Resource Group

    2) copied the xml files and checked syntax

    3) Deployed the policies on virtual nodes- however i  can see the policies enabled on both nodes. do I have to make any other changes to make it disabled on one?

    4) Also, while testing, I should be able to see the monitors enabled on only one physical node, right?

     

     

    Thanks,
    Sree

  • Hi 

  • Hello,

     

    Thanks for the repy. I have restarted the agent. Still  Monitors are active on both. 

     

    Thsi is from the active node

     

    Version Status
    --------------------------------------------------------------------
    CONFIGSETTINGS "For internal message filtering" enabled 0001.0002
    MONITOR "Processes-CME" enabled 0001.0001
    MONITOR "Processes-CMS" enabled 0001.0001
    MONITOR "SI-CPUSpikeCheck" enabled 0001.0000
    MONITOR "SI-GlobalCPUUtilization-AT" enabled 0001.0000
    MONITOR "SI-LinuxCifsUtilizationMonitor" enabled 0001.0000
    MONITOR "SI-LinuxNfsUtilizationMonitor" enabled 0001.0000
    MONITOR "SI-MemoryUsageAndPerformance" enabled 0001.0000
    MONITOR "SI-MemoryUtilization-AT" enabled 0001.0000
    MONITOR "SI-PerCPUUtilization-AT" enabled 0001.0000
    MONITOR "SI-RunQueueLengthMonitor-AT" enabled 0001.0000
    MONITOR "SI-SwapCapacityMonitor" enabled 0001.0000
    MONITOR "SI-SwapUtilization-AT" enabled 0001.0000
    monitortmpl "SI-CPUSpikeCheck" enabled 0001.0000
    monitortmpl "SI-GlobalCPUUtilization-AT" enabled 0001.0000
    monitortmpl "SI-LinuxCifsUtilizationMonitor" enabled 0001.0000
    monitortmpl "SI-LinuxNfsUtilizationMonitor" enabled 0001.0000
    monitortmpl "SI-MemoryUsageAndPerformance" enabled 0001.0000
    monitortmpl "SI-MemoryUtilization-AT" enabled 0001.0000
    monitortmpl "SI-PerCPUUtilization-AT" enabled 0001.0000
    monitortmpl "SI-RunQueueLengthMonitor-AT" enabled 0001.0000
    monitortmpl "SI-SwapCapacityMonitor" enabled 0001.0000
    monitortmpl "SI-SwapUtilization-AT" enabled 0001.0000

     

     

    From the other physical node

    * List installed policies for host 'localhost'.

    Version Status
    --------------------------------------------------------------------
    CONFIGSETTINGS "For internal message filtering" enabled 0001.0002
    mgrconf "HPOMandMA" enabled 0001.0002
    MONITOR "Processes-CME" enabled 0001.0001
    MONITOR "Processes-CMS" enabled 0001.0001
    OPCMSG "opcfilter" enabled 0001.0052
    OPCMSG "opcmsg" enabled 0011.0000

     

     

    Processes -CME and Processes CMS are the ones I created to be cluster aware

     

     

     

    Thanks,

    Sree

  • Can you try the attached files..

     

    I made a mistake in the apm file as I was in a hurry busy with other issues!! Sorry for that.

     

    Try ovappinstance -vc to verify the configuration.

     

    Execute the below commands:

     

    On itgsydc11n1 : ovconfchg -ns conf.cluster -set CLUSTER_LOCAL_NODENAME itgsydc11n1

     

    On itgsydc11n2 : ovconfchg -ns conf.cluster -set CLUSTER_LOCAL_NODENAME itgsydc11n2

     

    Restart the agent after that.

    APM.zip