Linux Cluster Node Status

Linux.ClusterMonitoring.NodeStatus.Monitor (UnitMonitor)

Monitor that alerts if a Cluster node does not report an Online status.

Element properties:

Source Code:

<UnitMonitor ID="Linux.ClusterMonitoring.NodeStatus.Monitor" Accessibility="Internal" Enabled="true" Target="Linux.ClusterMonitoring.Node" ParentMonitorID="Health!System.Health.AvailabilityState" Remotable="true" Priority="Normal" TypeID="UnixAuth!Unix.Authoring.ShellScript.MatchesRegExp.TwoState.MonitorType" ConfirmDelivery="false">

  <Category>AvailabilityHealth</Category>

  <AlertSettings AlertMessage="Linux.ClusterMonitoring.NodeStatus.Monitor.AlertMessage">

    <AlertOnState>Error</AlertOnState>

    <AutoResolve>true</AutoResolve>

    <AlertPriority>Normal</AlertPriority>

    <AlertSeverity>MatchMonitorHealth</AlertSeverity>

    <AlertParameters>

      <AlertParameter1>

              $Target/Host/Property[Type="Unix!Microsoft.Unix.Computer"]/NetworkName$

            </AlertParameter1>

      <AlertParameter2>$Data/Context///*[local-name()="StdOut"]$</AlertParameter2>

      <AlertParameter3>$Target/Property[Type="Linux.ClusterMonitoring.Node"]/ClusterName$</AlertParameter3>

    </AlertParameters>

  </AlertSettings>

  <OperationalStates>

    <OperationalState ID="StatusOK" MonitorTypeStateID="StatusOK" HealthState="Success"/>

    <OperationalState ID="StatusError" MonitorTypeStateID="StatusError" HealthState="Error"/>

  </OperationalStates>

  <Configuration>

    <Interval>300</Interval>

    <TargetSystem>

            $Target/Host/Property[Type="Unix!Microsoft.Unix.Computer"]/NetworkName$

          </TargetSystem>

    <ShellScript>

            

#!/bin/bash

sstat=`clustat`

startln=`echo "${sstat}" |grep -n ' Member Name'|awk '{print $1}' |cut -f1 -d:`

endln=`echo "${sstat}" |grep -n ' Service' |awk '{print $1}' |cut -f1 -d:`

startln=`expr $startln + 2`

endln=`expr $endln - 2`

nodestat=`printf "${sstat}" |sed -n ''${startln}','${endln}'p'`



anyOffline=`echo "${nodestat}"|grep Offline|wc -l`

if [ $anyOffline -ge 1 ]

then

        echo -e "ERROR:\n ${nodestat}"

else

        echo -e "OK:\n ${nodestat}"

fi



          </ShellScript>

    <ScriptArguments/>

    <Timeout>60</Timeout>

    <UserName>$RunAs[Name="Linux.ClusterMonitoring.Account"]/UserName$</UserName>

    <Password>$RunAs[Name="Linux.ClusterMonitoring.Account"]/Password$</Password>

    <ErrorRegExp>^ERROR</ErrorRegExp>

  </Configuration>

</UnitMonitor>