Edit

VMware ESX

Overview

VMWare is an software compagny based in USA. VMWare provides cloud computing and virtualization software and services.

The Centreon Plugin and Plugin-Packs rely on the Centreon VMWare Connector to request the vCenter API.

Plugin-Pack Assets

Monitored Objects

• ESX or ESXi

Discovery rules

Host

Services

Rule name	Description
Virt-VMWare2-Esx-HostDiscovery	Discover VMWare ESX linked to a vcenter or ESXi

Rule name	Description
Virt-VMWare2-Esx-Datastores-Latency-Name	Discover the Datastores and monitor latency
Virt-VMWare2-Esx-Nics-Traffic-Name	Discover the traffic NIC and monitor

Monitored Metrics

Metric name	Description	Unit
host.cpu.utilization.percentage	Total CPU usage in percentage	%
host.cpu.utilization.mhz	Total CPU usage in MHz	MHz
core#host.core.cpu.utilization.percentage	CPU usage per core in percentage	%

Metric name	Description	Unit
host.health.problems.current.count	Total number of problem on the ESXs	Count
host.health.problems.current.count	Number of health checks in green	Count
host.health.yellow.current.count	Number of health checks in yellow	Count
host.health.red.current.count	Number of health checks in red	Count

Metric name	Description	Unit
host.memory.usage.bytes	Memory used	Bytes
host.memory.overhead.bytes	Memory overhead needed by the VMs on the ESX	Bytes
host.memory.state.count	Memory state

Metric name	Description	Unit
Status	Overall status on the ESX

Metric name	Description	Unit
host.swap.in.usage.bytespersecond	Swap in rate Usage	B/s
host.swap.out.usage.bytespersecond	Swap out rate Usage	B/s

Metric name	Description	Unit
host.vm.poweredon.current.count	Number of powered on VMs on the ESX	Count
host.vm.poweredoff.current.count	Number of powered off VMs on the ESX	Count
host.vm.suspended.current.count	Number of suspended VMs on the ESX	Count

Prerequisites

Centreon VMWare Connector

For the VMWare monitoring, Centreon use daemon to connect and request the vCenter.

Install this daemon on each needed poller:

yum install centreon-plugin-Virtualization-VMWare-daemon

To configure the access to your infrastructure, edit the "/etc/centreon/centreon_vmware.pm" configuration file:

%centreon_vmware_config = (
    vsphere_server => {
        default => {
            url => 'https://<ip_hostname>/sdk',
            username => '<username>',
            password => '<password>'
        }
    }
);

1;

Make sure to replace variables with needed information:

• ip_hostname: IP address or hostname of the vCenter or ESX (if standalone),
• username: username with at least "read only" access to the vCenter or ESX (you can use domain user),
• password: password of the username.

You can configure multiple vCenter or ESXi connections using this structure:

%centreon_vmware_config = (
    vsphere_server => {
        'my_first_vcenter' => {
            url => 'https://<ip_hostname>/sdk',
            username => '<username>',
            password => '<password>'
        },
        'my_other_vcenter' => {
            url => 'https://<ip_hostname>/sdk',
            username => '<DOMAIN>\<username>',
            password => '<password>'
        },
    },
    port => 5700
);

1;

Each entry is called a container.

You can also define the "port" attribute to change listening port.

Then start the daemon and make sure it is configured to start at server boot:

systemctl start centreon_vmware
systemctl enable centreon_vmware

Make sure that the daemon configuration works fine by looking for errors in "/var/log/centreon/centreon_vmware.log".

Network flows

The Poller with the Centreon VMware Connector installed need to access in TCP/443 HTTPS to the vCenter.

The Pollers that request the Centreon VMWare Connector host need to access in TCP/5700 to the Centreon VMWare Connector host.

Installation

Online IMP Licence & IT-100 Editions

Offline IMP License

1. Install the Vmware2 Connector Centreon Plugin on every Poller expected to monitor the devices:

yum install centreon-plugin-Virtualization-Vmware2-Connector-Plugin

2. Install the 'Vmware ESX' Centreon Plugin-Pack on the "Configuration > Plugin packs > Manager" page

1. Install the Vmware ESX Centreon Plugin on every poller expected to monitor the devices:

yum install centreon-plugin-Virtualization-Vmware2-Connector-Plugin

2. Install the Centreon Plugin-Pack RPM:

yum install centreon-pack-virtualization-vmware2-esx

3. Install the 'Vmware ESX' Centreon Plugin-Pack on the "Configuration > Plugin packs > Manager" page

Configuration

• Log into Centreon and add a new Host through "Configuration > Hosts".
• Apply the relevant Host Template "Virt-VMWare2-ESX-custom", and configure the mandatory Macros:

Mandatory	Name	Description
X	CENTREONVMWARECONTAINER	Name of your container in the file centreon_vmware.pm
X	CENTREONVMWAREHOST	The Centreon server that launches the connection
X	CENTREONVMWAREPORT	By default: 5700
X	ESXNAME	Name of the ESX (defined in your VMWare infrastructure)
	CENTREONVMWAREEXTRAOPTIONS	Customize it with your own if needed

FAQ

How do I run my plugin through the CLI and what do the main parameters stand for ?

Once you've installed the plugin, you can test it logging with centreon-engine user:

/usr/lib/centreon/plugins//centreon_vmware_connector_client.pl
    --plugin=apps::vmware::connector::plugin
    --mode=cpu-host
    --custommode=connector
    --connector-hostname='localhost'
    --connector-port='5700'
    --container='vcenter01' 
    --esx-hostname='SRV-ESX-TLS'
    --unknown-status='%{status} !~ /^connected$/i'
    --warning-status=''
    --critical-status=''
    --warning-total-cpu='80'
    --critical-total-cpu='90'
    --warning-total-cpu-mhz=''
    --critical-total-cpu-mhz=''
    --warning-cpu=''
    --critical-cpu='' 

Expected command output is shown below:

OK: Host 'SRV-ESX-TLS' : status connected - cpu total average : 48.63 %, 14592.00 MHz - All CPUs are ok |
'cpu_total'=48.63%;;;0;100 'cpu_total_MHz'=14592.00MHz;;;0;30000 'cpu_0'=13.95%;;;0;100 'cpu_1'=12.01%;;;0;100
'cpu_2'=24.58%;;;0;100 'cpu_3'=24.55%;;;0;100 'cpu_4'=26.72%;;;0;100 'cpu_5'=24.38%;;;0;100 'cpu_6'=24.23%;;;0;100
'cpu_7'=26.37%;;;0;100 'cpu_8'=27.71%;;;0;100 'cpu_9'=26.16%;;;0;100

This command above checks the CPU Usage (--plugin=apps::vmware::connector::plugin --mode=cpu-host) of the ESX SRV-ESX-TLS (--esx-hostname='SRV-ESX-TLS'). It connects to the VMWare daemon on localhost (--connector-hostname='localhost') on the port 5700 (--connector-port='5700'). Then the command requests the container vcenter01 (--container='vcenter01') because the ESX SRV-ESX-TLS is managed by vcenter01.

It will trigger a WARNING alarm if the CPU Usage is above 80% (--warning-total-cpu='80') and a CRITICAL alarm if the CPU Usage is above 90% (--critical-total-cpu='90').

All available modes with the plugin can be displayed with:

/usr/lib/centreon/plugins/./centreon_vmware_connector_client.pl \
    --plugin=apps::vmware::connector::plugin \
    --list-mode

The available options for a mode can be displayed using the --help parameter:

/usr/lib/centreon/plugins/./centreon_vmware_connector_client.pl \
    --plugin=apps::vmware::connector::plugin \
    --mode=cpu-host \
    --help

Why do I get the following error:

UNKNOWN: Unknown container name 'default' |

This error message means that the container invoked in the command doesn't exist in your VMWare connector configuration. Check your macro CENTREONVMWARECONTAINER on your host or check the file /etc/centreon/centreon_vmware.pm

UNKNOWN: Container connection problem |

This error message means that you have a issue with the credentials provided for your Container. Check your credentials in /etc/centreon/centreon_vmware.pm. You can also take a look into the log for more information: /var/log/centreon/centreon_vmware.log

UNKNOWN: Cannot get value for counters (Maybe, object(s) cannot be reached: disconnected, not running, time not synced (see time-host mode),...) |

This error message means that you the plugin cannot get value for some counters. Most of the time it caused by a time shift between the Centreon Servers and the VMWare Infrastructure. Check the synchronization with the mode time-host.

UNKNOWN: Cannot get counter 'net.received.average' for the sampling period '300' (counter level: 2, sampling level: 1)

Some counters like 'mem.state.latest' and 'net.received.average' are only available with a sampling level at 2. You can configure the level sampling in the vCenter console for these counters. You can also configure the macro SERVICEEXTRAOPTIONS on the services where you catched this error with the option --sampling-period='20'.