Description

After a new MPC5E module was inserted, a 'warm temperature' alarm is displayed. 

Symptoms

A new chassis alarm is seen after a new MPC module was inserted into slot 2. From " show chassis hardware | match fpc ", it is a MPC5E module. Note that there are other types of MPC cards present in the chassis.

mx960> show chassis alarms | no-more
1 alarms currently active
Alarm time               Class  Description
2021-03-25 14:27:55 UTC  Minor  Temperature Warm

FPC 0            REV 41   750-028467   ABCV4836          MPC 3D 16x 10GE
FPC 1            REV 41   750-028467   CAEC4746          MPC 3D 16x 10GE
FPC 2            REV 77   750-045715   CANA5817          MPC5E 3D Q 24XGE+6XLGE
FPC 5            REV 18   750-045372   CADW4253          MPCE Type 3 3D
FPC 7            REV 41   750-028467   CACY3692          MPC 3D 16x 10GE
FPC 8            REV 20   750-038493   CAEE0301          MPCE Type 2 3D Q
FPC 9            REV 41   750-028467   ABCM8209          MPC 3D 16x 10GE


From the chassisd log, the chassis temperature is reported over 75 degrees C, but no fan/impeller failure is detected.

Mar 25 14:27:04  chassisd[3263]: %DAEMON-4-CHASSISD_HIGH_TEMP_CONDITION: Chassis temperature over 75 degrees C (but no fan/impeller failure detected)
Mar 25 14:27:04  alarmd[3672]: %DAEMON-4: Alarm set: Temp sensor color=YELLOW, class=CHASSIS, reason=Temperature Warm
Mar 25 14:27:04  craftd[3265]: %DAEMON-4: Minor alarm set, Temperature Warm
Mar 25 14:27:05  inetd[3270]: %DAEMON-4: /usr/sbin/sshd[33075]: exited, status 255
Mar 25 14:27:45  alarmd[3672]: %DAEMON-4: Alarm cleared: Temp sensor color=YELLOW, class=CHASSIS, reason=Temperature Warm
Mar 25 14:27:45  craftd[3265]: %DAEMON-4:Minor alarm cleared, Temperature Warm

Solution

This alarm may be trigged by the MPC5E module, which was confirmed from the temperature information on each component of the chassis.

mx960> show chassis environment no-forwarding
Class Item                           Status     Measurement
      FPC 2 PLX PCIe Switch TSe      OK         57 degrees C / 134 degrees F
      FPC 2 PLX PCIe Switch Chi      OK         76 degrees C / 168 degrees F

This is a temperature evaluation algorithm issue.

There was an algorithm change in Junos specific to MPC5E in terms of how temperatures are reported. For fixed Junos releases, refer to PR1199447 - An enhancement for implementing sensor-specific temperature thresholds is needed On MX240/480/960 and EX9200 platforms

 In the scenario above, the user was on Junos release 15.1F2-S20, which does not have the enhancement for MPC5E module.