Last update: November 29, 2024
check_esxi_hardware (formerly known as check_esx_wbem) is an open source monitoring plugin to monitor the hardware of ESXi (and previously ESX) servers. It queries the CIM (Common Information Model) server running on the ESXi server to retrieve the current status of all discovered hardware parts. The plugin can also be used as standalone script to check the hardware. The plugin is written in python and uses the pywbem module.
If you are looking for commercial support for this monitoring plugin, need customized modifications or in general customized monitoring plugins, contact us at Infiniroot.com.
167706 downloads so far...
Download plugin and save it in your Nagios/Monitoring plugin folder (usually /usr/lib/nagios/plugins, depends on your distribution). Afterwards adjust the permissions (usually chmod 755).
Community contributions welcome on GitHub repo.
20080820 Initial release by David Ligeret
20080821 Add verbose mode by David Ligeret
20090219 Add try/except to catch AuthError and CIMError by Joshua Daniel Franklin
20100202 Added HP Support (HealthState) by Branden Schneider
20100512 Combined different versions (Joshua and Branden) and added hardware type switch
20100628 Outputs server model, s/n and bios version and set Unknown as default exit code by Samir Ibradzic
20100702 GlobalStatus was incorrectly getting (re)set to OK with every CIM element check by Aaron Rogers
20100705 After last version all Dell servers return UNKNOWN instead of OK, added Aaron's logic for Dell checks as well
20101028 Changed text in Usage and Example so people dont forget to use https://
20110110
If Dell Blade Servers were used, Serial Number of Chassis instead of Blade was
returned - by Ludovic Hutin
20110207 Bugfix/new feature for Intel server systems by Carsten
Schoene
20110215
Plugin now catches Socket Error (Timeout Error) and added a timeout parameter
by Ludovic Hutin
20110221
Removed recently added timeout parameter due to incompatibility on Windows systems
20110221
Changed plugin name from check_esxi_wbem.py to check_esxi_hardware.py
20110426
Added 'ibm' hardware type (compatible to Dell output). Tested by Keith Erekson
on an IBM x3550
20110503
Plugin rewritten, added automatic hardware detection, opt params, perfdata and
much more by Phil
Randal
20110504
Some minor code changes, removed typo, bugfix for voltage sensors on IBM server
by Phil
Randal
20110505
Added possibility to use first line of a file as password (file:) by Fredrik
Åslund
20110507 A lot of bugfixes and enhancements from Phil
Randal (see changelog in plugin for details)
20110520
Bugfix for IBM Blade Servers by Bertrand Jomin
20110614
Rewrote external file handling, file can now be used for password AND username
20111003
Added ignore option to ignore certain elements by Ian
Chard
20120402
Making plugin GPL compatible (Copyright) and preparing for OpenBSD port
20120405
Fix lookup of warranty info for Dell by Phil
Randal
20120501
Bugfix in manufacturer discovery when cim entry not found or empty by Craig
Hart
20121027
Workaround for Dell PE x620 for Riser Config Err 0: Connected element (wrong
return code)
20130424
Another workaround for Dell systems "System Board 1 LCD Cable Pres 0: Connected"
20130702
Improving wrong authentication timeout and exit UNKNOWN by Carl R. Friend
20130725 Fix lookup of warranty info for Dell by Phil Randal
20140319 Another workaround for Dell systems "System Board 1 VGA Cable Pres 0: Connected"
20150109 Output serial number of chassis if a blade server is checked
20150119 Fix NoneType element bug by Andreas Gottwald
20150626 Added support for patched pywbem 0.7.0 and new version 0.8.0, handle SSL error exception
20150710 Exit Unknown instead of Critical for timeouts and auth errors by Stanislav German-Evtushenko
20151111 Cleanup and define variables by Stefan Roos
20160411 Distinguish between/add support for minor versions of pywbem 0.7 and 0.8
20160531 Add parameter for variable CIM port (useful when behind NAT)
20161013 Added support for pywbem 0.9.x (and upcoming releases)
20170905 Added option to ignore LCD/Display related elements (--no-lcd)
20180329 Try to use internal pywbem function to determine version
20180411 Throw an unknown if we can't fetch the data for some reason by Peter Newman
20181001 python3 compatibility
20190510 Allow regular expressions from ignore list (-r)
20190701 Fix lookup of warranty info for Dell (again) by Phil Randal
20200605 Added otion to ignore chassis intrusion elements (--no-intrusion) by Luca Berra
20200605 Add parameter (-S) for custom SSL/TLS protocol version
20200710 Improve missing mandatory parameter error text (issue #47), Delete temporary openssl config file after use (issue #48)
20210809 Fix TLSv1 usage (issue #51)
20220708 Added JSON-output (Zabbix needs it) by Marco Markgraf
20221230 Fix bug when missing S/N (issue #68)
20241129 Fix pkg_resources deprecation warning (issue #72), remove compatibility support for Python2 and pywbem 0.7.0
The FAQ have grown quite a lot and in order to support questions and comments, there is now a dedicated page: FAQ page.
To have a quick overview of the supported versions (ESXi, Python, PyWBEM), a dedicated page was created: check_esxi_hardware compatibility matrix.
The check_esxi_hardware plugin was initially built with support for legacy systems and very old Python and PyWBEM versions. But since November 2024, support for EOL versions has been removed. The requirements have accordingly split into two sections.
check_esxi_hardware.py uses the functions of the python module PyWBEM. It is therefore mandatory to install this module. Most distributions already offer pywbem as a package.
The Python way using pip (platform independent):
sudo pip install pywbem
sudo pip3 install pywbem
To upgrade pywbem to the latest version using pip (platform independent):
sudo pip install -U pywbem
sudo pip3 install -U pywbem
DEB based installation (Debian, Ubuntu, Linux Mint, ...):
Note that the pywbem package was removed in Debian >=11 and Ubuntu >=20.04, use pip/pip3 in this case.
sudo apt-get install python-pywbem
sudo apt-get install python3-pywbem
DNF based installation (RedHat, Rocky, Fedora, ...):
sudo dnf install python3-pywbem
Zypper based installation (SuSE):
sudo zypper install python-pywbem
Here are two guides how to install Python and PyWBEM on Windows and use the plugin as "standalone" script:
Short | Long | Description |
-H* | --host* | Hostname or IP address of ESX/ESXi server |
-U* | --user* | Username to check (must be a local user on target host) Note: If you don't want to use your root user, use this workaround to create a local ESXi user Use file:/path/to/.file to use first string as username |
-P* | --pass* | Password for given user Use file:/path/to/.file to use second string as password - this won't show the password in servers process list |
-C | --cimport | CIM port (default 5989) |
-S | --sslproto | Overwrite system default of SSL/TLS protocol to use. Must be one of: SSLv2, SSLv3, TLSv1.0, TLSv1.1, TLSv1.2, TLSv1.3 |
-V | --vendor | Define the vendor (hardware type) of the server: auto, dell, hp, ibm, intel, unknown (default) auto: Plugin tries to determine hardware itself by using CIM entries unknown: If no hw/vendor type was given, unknown will be used (like auto) |
-i | --ignore | Ignore given list (comma separated) of elements |
-r | --regex | Allow regular expression lookups of elements in ignore list |
-v | --verbose | Verbose/detailled output for debugging |
-p | --perfdata | Show performance data to create graphs (mainly temperature sensors and fan rpm) |
-I | --html | Add web-links to hardware manuals for Dell servers (use your country extension) |
-t | --timeout | Timeout in seconds Note: Some server models take a long time to display all results. Use this parameter accordingly. |
N/A | --format | Change output format. Values can be 'string' or 'json' (needed if you use Zabbix). Defaults to 'string'. |
N/A | --pretty | Changes the JSON output (--format json) to a human readable output. |
N/A | --no-power | Do not collect power performance data |
N/A | --no-volts | Do not collect voltage performance data |
N/A | --no-current | Do not collect current performance data |
N/A | --no-temp | Do not collect temperature performance data |
N/A | --no-fan | Do not collect fan performance data |
N/A | --no-lcd | Do not collect LCD/front display status data |
N/A | --no-intrusion | Do not collect chassic intrusion status data |
* Required parameters
Since version 20110505 it is possible to use a file as password-source. The string in the given file is used as password.
Since version 20110614 the file can be used for both username and password.
This enhances security! If no file is used, the username and password credentials will be shown in cleartext in the processlist when the plugin is executed.
Please watch out for the correct usage!
Example 1: You want to use a file (/home/nagios/.esxipass) which contains username and password. Note that two strings are separated by a space:
# cat /home/nagios/.esxipass
root mypass123
# ./check_esxi_hardware.py -H 172.17.16.131 -U file:/home/nagios/.esxipass -P file:/home/nagios/.esxipass -V dell
Example 2: You only want to use a file for the password. Note that there is only one string in the file:
# cat /home/nagios/.esxipass
mypass123
# ./check_esxi_hardware.py -H 172.17.16.131 -U root -P file:/home/nagios/.esxipass -V dell
Usage:
./check_esxi_hardware.py -H esxi-server-ip -U username
-P mypass [-C -S -V -i -r -v -p -I]
./check_esxi_hardware.py --host esxi-server-ip --user username --password mypass
[--cimport --sslproto --vendor --ignore --regex --verbose --perfdata --html]
./check_esxi_hardware.py -H esxi-server-ip -U -P file:/path/.passwdfile [--cimport --sslproto --vendor
--ignore --regex --verbose --perfdata --html]
./check_esxi_hardware.py -H esxi-server-ip -U file:/path/.passwdfile -P file:/path/.passwdfile
[-C -S -V -i -r -v -p --html]
Examples:
./check_esxi_hardware.py -H 10.0.0.200 -U root -P mypass
-V dell -p -I de
./check_esxi_hardware.py --host esxiserver1 --user root --password mypass --vendor
hp --perfdata
./check_esxi_hardware.py --host esxiserver2 --user root --password mypass --vendor
dell --html us
./check_esxi_hardware.py -H esxiserver1 -U root -P file:/root/.esxipass -V dell
./check_esxi_hardware.py -H esxiserver1 -U file:/root/.esxipass -P file:/root/.esxipass
-V dell
./check_esxi_hardware.py -H esxiserver1 -U root -P mypass -V dell -i "IPMI SEL"
./check_esxi_hardware.py -H esxiserver1 -U root -P mypass -C 5989 -S TLSv1.0
# 'check_esxi_hardware' command definition (basic)
define command{
command_name check_esxi_hardware
command_line $USER1$/check_esxi_hardware.py -H $HOSTADDRESS$ -U $ARG1$ -P $ARG2$ -V $ARG3$
}
# 'check_esxi_hardware' command definition (with appended optional parameter)
define command{
command_name check_esxi_hardware
command_line $USER1$/check_esxi_hardware.py -H $HOSTADDRESS$ -U $ARG1$ -P $ARG2$ -V $ARG3$ $ARG4$
}
The command definition in Icinga 2.x is already prepared because check_esxi_hardware is part of the ITL plugins. See https://www.icinga.com/docs/icinga2/latest/doc/10-icinga-template-library/#esxi_hardware. For the sake of completeness:
object CheckCommand "esxi_hardware" {
import "plugin-check-command"
command = [ PluginContribDir + "/check_esxi_hardware.py" ]
arguments = {
"-H" = {
value = "$esxi_hardware_host$"
description = "report on HOST"
}
"-U" = {
value = "$esxi_hardware_user$"
description = "user to connect as"
}
"-P" = {
value = "$esxi_hardware_pass$"
description = "password"
}
"-C" = {
value = "$esxi_hardware_port$"
description = "cim port"
}
"-S" = {
value = "$esxi_hardware_sslproto$"
description = "ssl/tls protocol"
}
"-V" = {
value = "$esxi_hardware_vendor$"
description = "Vendor code: auto, dell, hp, ibm, intel, or unknown"
}
"-I" = {
value = "$esxi_hardware_html$"
description = "generate html links for country XX"
}
"-i" = {
value = "$esxi_hardware_ignore$"
description = "comma-separated list of elements to ignore"
}
"-r" = {
set_if = "$esxi_hardware_regex$"
description = "Allow regular expression lookups of elements in ignore list"
}
"-p" = {
set_if = "$esxi_hardware_perfdata$"
description = "collect performance data for pnp4nagios"
}
"--format" = {
value = "$esxi_hardware_format$"
description = "Set output format to string or json (defaults to string)"
}
"--pretty" = {
set_if = "$esxi_hardware_pretty$"
description = "Show plugin output in a human readable format (in combination with --format json)"
}
"--no-power" = {
set_if = "$esxi_hardware_nopower$"
description = "don't collect power performance data"
}
"--no-volts" = {
set_if = "$esxi_hardware_novolts$"
description = "don't collect voltage performance data"
}
"--no-current" = {
set_if = "$esxi_hardware_nocurrent$"
description = "don't collect current performance data"
}
"--no-temp" = {
set_if = "$esxi_hardware_notemp$"
description = "don't collect temperature performance data"
}
"--no-fan" = {
set_if = "$esxi_hardware_nofan$"
description = "don't collect fan performance data"
}
"--no-lcd" = {
set_if = "$esxi_hardware_nolcd$"
description = "don't collect lcd/front display status data"
}
"--no-intrusion" = {
set_if = "$esxi_hardware_nointrusion$"
description = "don't collect chassic intrusion status data"
}
}
vars.esxi_hardware_host = "$address$"
vars.esxi_hardware_port = 5989
vars.esxi_hardware_regex = false
vars.esxi_hardware_perfdata = false
vars.esxi_hardware_nopower = false
vars.esxi_hardware_novolts = false
vars.esxi_hardware_nocurrent = false
vars.esxi_hardware_notemp = false
vars.esxi_hardware_nofan = false
vars.esxi_hardware_nolcd = false
vars.esxi_hardware_nointrusion = false
}
Basic check on a HP server:
# Check HP Server hardware
define service{
use generic-service
host_name esxi1
service_description Hardware
check_command check_esxi_hardware!root!mypass!hp
}
Service check on a DELL Server with perfdata:
# Check DELL Server hardware
define service{
use generic-service
host_name esxi2
service_description Hardware
check_command check_esxi_hardware!root!mypass!dell!--perfdata
}
>Service check on a HP Server with user and password read from file:
# Check HP Server hardware
define service{
use generic-service
host_name esxi1
service_description Hardware
check_command check_esxi_hardware!/home/nagios/.esxipass!/home/nagios/.esxipass!hp
}
Service check on an IBM Server where System Event Log alerts should be ignored:
# Check IBM Server hardware
define service{
use generic-service
host_name esxi3
service_description Hardware
check_command check_esxi_hardware!root!mypass!ibm!-i "IPMI SEL"
}
The following example is a standard service object checking a Dell server with perfdata enabled and using a file for the password:
# Hardware Check
object Service "Hardware" {
import "generic-service"
host_name "myesxiserver1"
check_command = "esxi_hardware"
vars.esxi_pass = "file:/var/lib/nagios/.esxipass"
vars.esxi_vendor = "dell"
}
The next example is more interesting as it uses apply rules (a feature of Icinga 2.x). Let's assume we have a couple of ESXi hosts already defined with custom attributes:
object Host "myesxiserver1" {
import "generic-host"
address = "10.50.8.71"
vars.os = "ESXi"
vars.server.vendor = "Cisco"
}
Using these custom attributes (vars.os and vars.server.vendor) we can now create an apply rule:
# Hardware Checks of Cisco UCS ESX servers
apply Service "Hardware" {
import "generic-service"
check_command = "esxi_hardware"
vars.esxi_hardware_user = "root"
vars.esxi_hardware_pass = "file:/var/lib/nagios/.esxipass"
vars.esxi_hardware_vendor = "auto"
vars.esxi_hardware_perfdata = true
assign where host.address && host.vars.os == "ESXi" && host.vars.server.vendor == "Cisco"
}
This "Hardware" service object will now be applied to all hosts which have the custom host attributes host.vars.os set to "ESXi" and host.vars.server.vendor set to "Cisco".