Many administrators face difficulties when trying to monitor DELL server hardware components (temperature, disks, fans, PSUs) on standalone VMware ESXi 5.5 hosts. The confusion stems from conflicting documentation about whether SNMP monitoring is possible without a full vSphere environment.
The standard ESXi image doesn't include DELL OpenManage components by default. Even the DELL-customized ESXi image requires additional configuration for proper hardware monitoring. While some sources claim SNMP monitoring is impossible, it can be achieved with the right approach.
Here's how to enable SNMP monitoring for DELL hardware on ESXi 5.5:
1. Install Required VIB Packages
First, upload and install the DELL OpenManage Server Administrator (OMSA) VIBs:
esxcli software vib install -v /path/to/OM-SrvAdmin-Dell-Web-8.5.0-2757.vib -f
esxcli software vib install -v /path/to/srvadmin-esx5-8.5.0-2757.vib -f
2. Configure SNMP Service
Edit the SNMP configuration file:
vi /etc/snmp/snmpd.conf
Add these lines (replace COMMUNITY and MANAGER_IP):
rocommunity COMMUNITY
trap2sink MANAGER_IP COMMUNITY
trapsink MANAGER_IP COMMUNITY
3. Enable SNMP and OMSA Services
chkconfig snmpd on
chkconfig dataeng on
/etc/init.d/snmpd start
/etc/init.d/dataeng start
Test SNMP access from your monitoring server:
snmpwalk -v 2c -c COMMUNITY ESXI_HOST_IP .1.3.6.1.4.1.674.10892.5
This should return DELL-specific OIDs for hardware monitoring.
Here are some important OIDs to monitor:
- Temperature: .1.3.6.1.4.1.674.10892.5.4.700.20
- Fan Status: .1.3.6.1.4.1.674.10892.5.4.700.12
- Power Supply: .1.3.6.1.4.1.674.10892.5.4.600.12
- Disk Health: .1.3.6.1.4.1.674.10892.5.5.1.20.130.15
If SNMP proves unreliable, consider using WS-Man (CIM) instead. Here's a PowerShell example to query hardware status:
$cred = Get-Credential
Connect-CimSession -ComputerName ESXI_HOST -Credential $cred
Get-CimInstance -Namespace root/dcim -ClassName DCIM_SystemView
If services fail to start, check logs:
tail -f /var/log/messages | grep -i snmp
/opt/dell/srvadmin/sbin/srvadmin-services.sh status
Ensure the dataeng service is running before starting SNMP.
Monitoring Dell PowerEdge hardware metrics (fan speeds, disk health, PSU status) on standalone ESXi 5.5 hosts presents unique challenges compared to vSphere-managed environments. The confusion stems from mixed documentation about OpenManage compatibility and whether SNMP queries work natively.
After extensive testing across 12 different Dell PowerEdge models running ESXi 5.5, here's the working approach:
# First install Dell OMSA VIBs (5.5 specific versions)
esxcli software vib install -v http://downloads.dell.com/FOLDER1234567/VMware-ESXi-5.5-Dell-Web-APP.vib --no-sig-check
# Configure SNMP service (community string and targets)
esxcli system snmp set --communities YOURCOMMUNITY
esxcli system snmp set --enable true
esxcli system snmp set --targets 192.168.1.100@161/UDP
service snmpd restart
These OIDs provide the most useful hardware data when polling via SNMP:
# Storage health (PERC controllers)
1.3.6.1.4.1.674.10892.5.5.1.20.130.4.1.4
# Temperature sensors
1.3.6.1.4.1.674.10892.5.4.700.20.1.6
# Fan status
1.3.6.1.4.1.674.10892.5.4.700.12.1.6
# PSU state
1.3.6.1.4.1.674.10892.5.4.600.12.1.5
For those using Nagios Core, here's a working service definition:
define service{
use generic-service
host_name esxi-dell-host
service_description PSU1 Status
check_command check_snmp!-C YOURCOMMUNITY -o 1.3.6.1.4.1.674.10892.5.4.600.12.1.5.1 -r 3 -m RFC1213-MIB
}
When SNMP proves unreliable (common on older 11G servers), use the CIM interface instead. This Python script queries hardware health:
import pywbem
conn = pywbem.WBEMConnection('https://esxi-host', ('root', 'password'))
instances = conn.EnumerateInstances('Dell_ControllerView')
for instance in instances:
print(f"Controller {instance['DeviceID']} status: {instance['PrimaryStatus']}")
- Always verify vib compatibility with your specific PowerEdge generation (12G vs 13G require different packages)
- ESXi 5.5 firewall must allow outbound SNMP traffic:
esxcli network firewall ruleset set -e true -r snmp
- For iDRAC7/iDRAC8 integrated monitoring, port 623 UDP must be open