Monitor Vmware ESX

[GigaGuy]

Does anyone know if there are any plugins like the ones at the bottom of this page? http://www.zenoss.com/product/screenshots for monitoring VM's

[karias]

I'm also running on v 8.7a and I got this going. The only thing I've noticed is the cpu graphs for majority of the vmware servers look like it's graphing at 100% instead of the correct cpu utilization. Only a handful are graphing correctly. Any one else experiencing this?


*Update - Never mind....it looks like it's going over 100%.

[fmangeant]

Hi

could you please post pictures as png / jpeg ?

Thanks in advance !

[warinetsa]

I'm also running on v 8.7a and I got this going. The only thing I've noticed is the cpu graphs for majority of the vmware servers look like it's graphing at 100% instead of the correct cpu utilization. Only a handful are graphing correctly. Any one else experiencing this?

Hi there,

I experienced the same problem with ESX 3.0.1. It disappeard with 3.0.2.

Cheers
Manfredo

[mattp]

Hi,

I imported the ESX template successfully and added one of our ESX hosts. SNMP is configured on it and snmpwalk retrieves data successfully:

Code: Select all

# snmpwalk -v2c -c blabla myhost .1.3.6.1.4.1.6876.2.1.1.1
SNMPv2-SMI::enterprises.6876.2.1.1.1.0 = INTEGER: 0
SNMPv2-SMI::enterprises.6876.2.1.1.1.1 = INTEGER: 1
SNMPv2-SMI::enterprises.6876.2.1.1.1.2 = INTEGER: 2
SNMPv2-SMI::enterprises.6876.2.1.1.1.3 = INTEGER: 3
SNMPv2-SMI::enterprises.6876.2.1.1.1.4 = INTEGER: 4
SNMPv2-SMI::enterprises.6876.2.1.1.1.5 = INTEGER: 5
SNMPv2-SMI::enterprises.6876.2.1.1.1.6 = INTEGER: 6
SNMPv2-SMI::enterprises.6876.2.1.1.1.7 = INTEGER: 7
SNMPv2-SMI::enterprises.6876.2.1.1.1.8 = INTEGER: 8
SNMPv2-SMI::enterprises.6876.2.1.1.1.9 = INTEGER: 9
SNMPv2-SMI::enterprises.6876.2.1.1.1.10 = INTEGER: 10
SNMPv2-SMI::enterprises.6876.2.1.1.1.11 = INTEGER: 11
SNMPv2-SMI::enterprises.6876.2.1.1.1.12 = INTEGER: 12
SNMPv2-SMI::enterprises.6876.2.1.1.1.13 = INTEGER: 13
SNMPv2-SMI::enterprises.6876.2.1.1.1.14 = INTEGER: 14
SNMPv2-SMI::enterprises.6876.2.1.1.1.15 = INTEGER: 15

But the problem is when I add the ESX Host inside cacti, there are no items listed for the data queries ( see the attached picture ).

Verbose Query works and tells the following:

Code: Select all

+ Running data query [15].
+ Found type = '6 '[script query].
+ Found data query XML file at '/var/www/html/cacti/resource/script_server/vmware_mem.xml'
+ XML file parsed ok.
+ Executing script for list of indexes '/usr/bin/php -q /var/www/html/cacti/scripts/ss_vmware_mem.php 10.1.1.201 9 2:161:500:nagios:::MD5::[None] index'
+ Executing script query '/usr/bin/php -q /var/www/html/cacti/scripts/ss_vmware_mem.php 10.1.1.201 9 2:161:500:nagios:::MD5::[None] query index'
+ Executing script query '/usr/bin/php -q /var/www/html/cacti/scripts/ss_vmware_mem.php 10.1.1.201 9 2:161:500:nagios:::MD5::[None] query displayname'
+ Executing script query '/usr/bin/php -q /var/www/html/cacti/scripts/ss_vmware_mem.php 10.1.1.201 9 2:161:500:nagios:::MD5::[None] query vmid'
+ Executing script query '/usr/bin/php -q /var/www/html/cacti/scripts/ss_vmware_mem.php 10.1.1.201 9 2:161:500:nagios:::MD5::[None] query gueststate'
+ Executing script query '/usr/bin/php -q /var/www/html/cacti/scripts/ss_vmware_mem.php 10.1.1.201 9 2:161:500:nagios:::MD5::[None] query vmstate'
+ Executing script query '/usr/bin/php -q /var/www/html/cacti/scripts/ss_vmware_mem.php 10.1.1.201 9 2:161:500:nagios:::MD5::[None] query total'
+ Found data query XML file at '/var/www/html/cacti/resource/script_server/vmware_mem.xml'
+ Found data query XML file at '/var/www/html/cacti/resource/script_server/vmware_mem.xml'
+ Found data query XML file at '/var/www/html/cacti/resource/script_server/vmware_mem.xml'

I don't know where the problem could be.

Hope that somebody can help me with this problem!

Thanks,
Matthias

[karias]

mattp, what version of cacti are you running?

[mattp]

mattp, what version of cacti are you running?

The latest stable, 0.8.7a.

[karias]

Have you updated your php scripts?

include_once(dirname(__FILE__) . "/../include/global.php");
instead of
include_once(dirname(__FILE__) . "/../include/config.php");

[nochance]

I am having the same problem with the CPU Graphs. I did update my scripts and I checked them twice to make sure the config.php was changed to global.php.

Matt, are you using Ping or SNMP to verify your servers are up? I had to switch from Ping to SNMP and that fixed the problem.

Also, it appears that ESX3 has removed the VMWare specific SNMP subagents that allow you to pull data from the VM's. This seems to be a big part of the reason a lot of us are having problems.

You can use the SDK and pull the data directly from the VCMS server or directly from an ESX server.

More details: http://unnoc.org/

Maybe someone would like to take a stab at converting the VI Perl Toolkit calls to work with Cacti? http://sourceforge.net/projects/viperltoolkit/

[nochance]

Also, I am trying to track users logged in to get an idea of how many people are using the VM's. It appears to be incrementing instead of clearing the data .

Example, if I log in today, then log again tomorrow, it shows 2 users logged in. If I log in the following day, it shows 3 users logged in. Anything we can do to fix that?


DOH! It's a Unix tool and I'm using it on Windows. Wonder if I can fix that somehow....

[apperrault]

Hi all,
I have a question. How can i see the Data Stores that my ESX server is using? I have 3 146GB data stores, and none of them show up in my graphs. I have tried the following Data queries and they all show me 0 rows:

Unix - Get Mounted Partitions
ucd/net - Get Monitored Partitions

SNMP - Get Mounted Partitions returns 6 rows, which are my partitions that ESX is installed on, but i see nothing for monitoring the Datastore drives.

Anyone have any thoughts? I can send an SNMP walk if that would help

app

[ymartin59]

Hello,
According to VMware MIB, the CPU utilization is "Time the virtual machine has been running on the CPU (seconds)."
So the CDEF "CPU Seconds As Percentage" should be something like "cdef=CURRENT_DATA_SOURCE,300,/"
But resulting graph are "wrong" compared to esxtop output. I guess the number of hardware CPU should be considered to get a relevant percentage.
On two servers, it seems "/ 10" give a percentage closed to the reality, but why ??

Another improvement point: network counters has been deprecated and new 64 bits counters (High Capacity) has been defined. See /usr/lib/vmware/snmp/mibs/VMWARE-RESOURCES-MIB.mib for details.

Yves

[ymartin59]

SNMP - Get Mounted Partitions returns 6 rows, which are my partitions that ESX is installed on, but i see nothing for monitoring the Datastore drives.
Anyone have any thoughts? I can send an SNMP walk if that would help

vmfs partitions are not visible in the service console except with vdf - I guess it comes from a file in /proc/vmware/
These partitions are not published in the vmware MIB either (see content in /usr/lib/vmware/snmp/mibs)
So, I would answer - no, it is not possible to monitor easily the datastores through SNMP.

[cynicismic]

Has anyone got round to updating this for ESX 3.5 yet?
It does work for the base ESX server but monitoring for each of the VM's doesn't work - looks like the mib's have changed.

Might get round to doing it myself if no-one else already has..

[cc]

I have ESX 3.5.0 and cacti version 0.8.7a and it works well !

Code: Select all

# cat /etc/snmp/snmpd.conf
$# Sample snmpd.conf containing VMware MIB module entries.

# This is a simple snmpd.conf that may help you test SNMP.
# It is not recommended for production use. Consult the
# snmpd.conf(5) man pages to set up a secure installation.

syscontact root@localhost (edit snmpd.conf)
syslocation room1 (edit snmpd.conf)
rocommunity public
trapcommunity public
trapsink localhost
trapsink 192.168.1.200

# VMware MIB modules. To enable/disable VMware MIB items
# add/remove the following entries.
dlmod SNMPESX            /usr/lib/vmware/snmp/libSNMPESX.so

# Allow Systems Management Data Engine SNMP to connect to snmpd using SMUX
smuxpeer .1.3.6.1.4.2.675.10824.1

and on cacti: