Troubleshooting "Fan Not Detected" Alarm on Versa Appliances

Category: Hardware / Alarms   |   Severity: Major

Overview

The "Fan Not Detected" alarm indicates that one or more cooling fans in the appliance are not being detected by the baseboard management controller (BMC) or the operating system. This can be caused by a failed fan, loose connector, firmware issue, or sensor reporting anomaly.

⚠️ Left unresolved, this condition risks thermal damage to the appliance.

Symptoms

• Alarm raised in Versa Director: "Fan Not Detected" (Major or Critical)
• Fan status shows ns (not present / no reading) in sensor output
• Director health dashboard shows hardware fault for the affected device

Diagnostic Commands

1. Check Raw Sensor Readings — Linux shell on the appliance

$ sudo ipmitool sensor

Lists all IPMI sensors with their current readings, units, status, and thresholds. Look for fan entries.

Example — healthy:

FAN1             | 3600.000   | RPM  | ok    | 500.000   | ...
FAN2             | 3600.000   | RPM  | ok    | 500.000   | ...
FAN3             | 3600.000   | RPM  | ok    | 500.000   | ...

Example — fan not detected:

FAN1             | na         |      | ns    | na        | na | na | na | na | na
FAN2             | 3600.000   | RPM  | ok    | 500.000   | ...

Key status codes:

2. Check SDR (Sensor Data Record) Table — Linux shell on the appliance

$ sudo ipmitool sdr

Lists all sensors registered in the SDR repository with their current value and status. More concise than sensor output.

Example — fan not detected:

FAN1             | na         | ns
FAN2             | 3600 RPM   | ok
FAN3             | 3600 RPM   | ok

To filter only fan entries:

$ sudo ipmitool sdr | grep -i fan

To get a detailed SDR entry for a specific fan:

$ sudo ipmitool sdr get "FAN1"

3. Check from Versa CLI — Director or Device CLI

admin@SanJose-DC-Standby-cli> show device sensors

Queries hardware sensor data through the Versa VOS layer, retrieving BMC/IPMI data in Versa CLI format.

Example — fan not detected:

Sensor Name      | Reading     | Unit | Status
-----------------+-------------+------+------------
FAN1             | N/A         | RPM  | Not Present
FAN2             | 3600        | RPM  | OK
FAN3             | 3600        | RPM  | OK
CPU Temp         | 42          | C    | OK
Inlet Temp       | 28          | C    | OK

A status of Not Present or N/A for a fan confirms the alarm.

Triage Steps

1. Identify which fan(s) are affected

   Run all three commands above and note which FAN# shows ns / Not Present.

2. Physical inspection (if site access is available)

   • Power down the appliance safely.
   • Reseat the fan module — loose connectors are a common cause.
   • Check for visible damage, dust blockage, or bent pins.
   • Power back on and re-run the sensor commands.

3. BMC cold reset (if physical access is not possible)

   $ sudo ipmitool mc reset cold

   Wait ~60 seconds, then re-check:

   $ sudo ipmitool sdr | grep -i fan

   A cold MC reset reloads the BMC firmware and may recover transient sensor reporting errors.

4. Check BMC firmware and IPMI connectivity

   $ sudo ipmitool mc info

   Verify BMC firmware version is current. Outdated BMC firmware can cause false sensor absence reports.

5. Check system event log for related entries

   $ sudo ipmitool sel list | grep -i fan

   This may surface historical fan failure events or sensor errors.

Resolution

Clearing the Alarm in Director

Once the underlying issue is resolved and sensor readings return to ok:

1. Verify the fix with:

   $ sudo ipmitool sdr | grep -i fan
   admin@<device>-cli> show device sensors

2. The alarm should auto-clear once Director polls and receives a healthy sensor status.

3. If the alarm persists after the hardware fix:

   • Navigate to Director → Monitor → Alarms
   • Locate the Fan Not Detected alarm for the device
   • Select → Clear Alarm (if auto-clear does not trigger within 5 minutes)

Escalation to Versa TAC

If the fan is physically replaced and the alarm persists, or if multiple fans show ns simultaneously on a healthy unit, escalate to Versa TAC with the following outputs:

• sudo ipmitool sensor
• sudo ipmitool sdr
• sudo ipmitool sel list
• sudo ipmitool mc info
• Director alarm details (device name, alarm time, alarm ID)