Fetching IPMI sensor data hangs a lot

Hi everyone,

Did someone else encountered this issue after the latest updates ?

Since latest updates when changes were done at IPMI level, LibreNMS started to work very hard, spiking the CPU very often above 70%. Polling time increased for each Dell Device that has IPMI enabled.

Tried to work with the values from the IPMI Settings, but nothing works.

On this execution:

Fetching IPMI sensor data…SNMP[‘/usr/bin/ipmitool’ ‘-H’ ‘HOSTNAME’ ‘-U’ ‘USER’ ‘-P’ ‘PASSWORD’ ‘-L’ ‘USER’ ‘-p’ ‘623’ ‘-N’ ‘3’ ‘-I’ ‘lanplus’ ‘-c’ ‘COMMUNITY’]

it hangs quite a lot of time and I’m assuming is the cause of the problem.

Thank you!

These are the seconds of the polling with IPMI enabled:

device1 2025-11-26 10:13:03 General Unknown 178.59
device2 2025-11-26 10:12:27 General Unknown 143.19
device3 2025-11-26 10:13:44 General Unknown (edit /etc/snmp/snmpd.conf) 16.93
device4 2025-11-26 10:14:00 General Unknown (edit /etc/snmp/snmpd.conf) 16.59

All 4 devices are Dell R630 with iDrac 8, Lifecycle Firmware Version 2.86.86.86.

Did you try to run snmpwalk pointed at one of the affected machines to see if it shows you anything weird?

I’ve done a manual ipmitool command on both iDrac8 and iDrac9:

iDRAC8: time ipmitool -I lanplus -C 3 -H 192.168.1.1 -U test -P test sdr list
real 2m34.308s
user 0m0.156s
sys 0m0.105s

iDRAC9: time ipmitool -I lanplus -C 3 -H 192.168.1.12 -U test -P test sdr list
real 2m53.269s
user 0m0.193s
sys 0m0.151s

So, it’s hard to believe both are broken. It’s the way it works, IPMI is slow.

What I want to understand is what changed and if there is a fix.

Thank you!

Hi,

All Dells are affected. It’s not something isolated.

I’m not seeing anything out of the normality. With IPMI disabled, everything runs relatively fast.

What do you mean by weird ?

So your manual test of just running ipmitool is taking between 2 and 3 minutes. If that’s how slow the call is then we can’t speed that up.

Than, what’s the explanation that before latest updates on the IPMI side, never got this issue ? what actually changed ?

I understand it takes 2-3 minutes, all 30+ Dells in the monitoring have that time, iDrac8 or 9. It makes me think that’s their implementation.

It might help me to understand if it’s truly a problem on my side, if others could help with some values on the response time.

Thank you!

On SuperMicro servers there was similar problem with IPMI

  • on one X11 helps disable HR Module ( Librenems > Device > Edit > Module > toggle HR).
  • on second helps Bios/IPMI Upgrade