Cheshire Cat Computing

Hi,
first of all sorry for my bad english, but i will try my best.

I have some problems monitoring the ESX Server with check_esx2.

ESX Version: 3.0
check_esx2: 2.4

Ok, snmp is working and i can connect via snmp from the nagios machine.

If I tryed following

check_esx2 -N -H ESX-SERVER-IP -l LIST

I recieved this output:

VHosts: 2/2 up: VHOST01(736), VHOST02(656)

Looks good

Then I tryed this

check_esx2 -N -H ESX-SERVER-IP -l MEM -v VHOST01 -w 20 -c 10
Please wait, data being gathered.
or
check_esx2 -N -H ESX-SERVER-IP -l CPU -v VHOST01 -w 90 -c 95
Gathering statistics, please wait.

Every time i recieved the same answer

Ok, I tryed to understand how the plugin is working, but i'm not a perl guru

Found that the vmids on the proc-filesystem (/proc/vmware/vm/*/names) differ from the one given by snmp (oid: .1.3.6.1.4.1.6876.2.1.1)

Info from proc:

for i in `find /proc/vmware/vm -name names`; do cat $i | awk '{print $1" "$19}'; done
vmid=1216 displayName="VHOST01"
vmid=1208 displayName="VHOST02"

from snmp:
snmpwalk -v1 -c public ESX-SERVER-IP .1.3.6.1.4.1.6876.2.1.1

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.6876.2.1.1.2.0 = STRING: "VHOST01"
SNMPv2-SMI::enterprises.6876.2.1.1.2.1 = STRING: "VHOST02"
SNMPv2-SMI::enterprises.6876.2.1.1.3.0 = STRING: "/vmfs/volumes/44e47ade-ec3b5cc8-dc87-000423b9b15d/VHOST01/VHOST01.vmx"
SNMPv2-SMI::enterprises.6876.2.1.1.3.1 = STRING: "/vmfs/volumes/44e47ade-ec3b5cc8-dc87-000423b9b15d/VHOST02/VHOST02.vmx"
SNMPv2-SMI::enterprises.6876.2.1.1.4.0 = STRING: "Microsoft Windows Server 2003, Standard Edition"
SNMPv2-SMI::enterprises.6876.2.1.1.4.1 = STRING: "Microsoft Windows Server 2003, Standard Edition"
SNMPv2-SMI::enterprises.6876.2.1.1.5.0 = INTEGER: 1000
SNMPv2-SMI::enterprises.6876.2.1.1.5.1 = INTEGER: 1500
SNMPv2-SMI::enterprises.6876.2.1.1.6.0 = STRING: "poweredOn"
SNMPv2-SMI::enterprises.6876.2.1.1.6.1 = STRING: "poweredOn"
SNMPv2-SMI::enterprises.6876.2.1.1.7.0 = INTEGER: 736 <== vmVMID
SNMPv2-SMI::enterprises.6876.2.1.1.7.1 = INTEGER: 656 <== vmVMID
SNMPv2-SMI::enterprises.6876.2.1.1.8.0 = STRING: "running"
SNMPv2-SMI::enterprises.6876.2.1.1.8.1 = STRING: "running"

In /var/tmp/vmware-stats.log i can see the vmid from the proc-filesystem.

Shouldn't the VMIDS be the same? (proc and snmp)

Someone any idea?

Thanks

I would have expected them to be the same... the plugin uses the SNMP ones in preference.

You can test the data retrieval by running the /etc/snmp/vmware-stats script from the command line on the ESX server. The first run will not output much data, but run it again after a couple of minutes and it will have enough history to give CPU percentages.

I dont have an ESX 3 machine here yet, so I can't test under this.

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11