Nagios: monitor Hyper-V 2012 health – BPA & VM

This Nagios plugin check Hyper-V 2012 server health through PowerShell + NRPE (nsclient++):

• Execute and retrieve BPA scan result (Best Practice Analyzer)
• Check that all VM are running
• Check that no VM are in any state other than “Operating normally”

Warning and Critical state are triggered by the sum of:

• BPA alerts
• # VM in unhealthy state

VM (models…) can be excluded through an argument

A new BPA scan is triggered at every execution, but new result is retrieved next time (async) (take too much time).

Provided performance data:

• # VM stopped,
• # VM unhealthy,
• # BPA errors
• # BPA warning

Tested setup

Linux:

• Centos 6.4 x64
• Nagios 3.4.4
• check_nrpe 2.13
• Centreon 2.4.2

Windows:

• Windows Server 2012 (needed to have PowerShell Hyper-V module)
• nsclient++ 0.4.1 x64

Script argument

• excludeVM (comma separated)
• maxWarn (Warning if aboce) (1 by default)
• maxError (Criticial if above) (5 by default)

maxWarn and maxCrit must be integer

Usages samples

Directly in PowerShell:

PS C:Program FilesNSClient++scripts>. .lotp_check_hyper-v.ps1
WARNING: 'VM Not Running:'2 'VM issues:'0 BPA Error:1 BPA Warning:12|VM_Not_Running=2 VM_issues=0 BPA_Error=1 BPA_Warning=12
PS C:Program FilesNSClient++scripts>

Through NRPE:

[root~]# /usr/lib64/nagios/plugins/check_nrpe -H myMonitoredServer -n -c check_hyper-v
WARNING: 'VM Not Running:'2 'VM issues:'0 BPA Error:1 BPA Warning:12|VM_Not_Running=2 VM_issues=0 BPA_Error=1 BPA_Warning=12
[root~]#

Install:

On Windows:

• Enable powershell script execution without signed : Set-ExecutionPolicy RemoteSigned
• copy script in folder C:Program FilesNSClient++scripts
• Add to nsclient.ini:
  • [/settings/external scripts/wrapped scripts]
  • check_hyper-v=lotp_check_hyper-v.ps1 -excludeVM $ARG1$ -maxWarn $ARG2$ -maxError $ARG3$

Setup:

On Centreon, by adding this command:

$USER1$/check_nrpe -H $HOSTADDRESS$ -n -t 60 -c check_hyper-v -a $ARG1$ $ARG2$ $ARG3$

Download

lotp_check_hyper-v.ps1

Source code:

# ====================================================================
# Ping a list of targets through NRPE
# Author: Mathieu Chateau - LOTP
# mail: mathieu.chateau@lotp.fr
# version 0.1
# ====================================================================

#
# Require Set-ExecutionPolicy RemoteSigned.. or sign this script with your PKI 
#

# ============================================================
#
#  Do not change anything behind that line!
#
param 
(
	[string]$targets,
	[int]$maxWarn = 1,
	[int]$maxError = 5
)

$output=""
$exitcode=2
$countOK=0
$countKO=0
$targetsArray=@()
$targetsArray=$targets -split(' ')
Remove-Job -Name * -Confirm:$false -Force
foreach($t in $targetsArray)
{
	Start-Job -Name $t -ArgumentList $t -ScriptBlock {param($t);if(Test-Connection -ComputerName $t  -Count 2 -Quiet -ErrorAction SilentlyContinue){return $true}else{return $false}} |Out-Null
}
while(Get-Job -State Running)
{
	Start-Sleep -Milliseconds 500
}
foreach ($job in Get-Job)
{
	$temp=Receive-Job -Name $job.Name
	if($temp)
	{
		$countOK++
	}
	else
	{
		$countKO++
		$output+=$job.Name+" - "
	}
}
if ($countKO -gt $maxError)
{
	$state="CRITICAL"
	$exitcode=2
}
elseif ($countKO -gt $maxWarn)
{
	$state="WARNING"
	$exitcode=1
}
else
{
	$state="OK"
	$exitcode=0
}

$output=$state+":"+$countOK+" online"+" - "+$countKO+" offline - "+$output
$output+='|'
$output+="online="+$countOK+";"+$maxWarn+";"+$maxError+";"+" "
$output+="offline="+$countKO+";"+$maxWarn+";"+$maxError+";"
Write-Host $output
exit $exitcode