I've got a pair of EX2200-48Ts that were recently upgraded to JUNOS 12.3R12.4 and combined into a Virtual Chassis. After just a few days uptime, the VC starts reporting a «FPC 1 CPU Temp Sensor Too Warm» yellow alarm most of the time. The FPC 1 switch was previously running JUNOS 11.4R7.5 for years and did never show this errror, and I've confirmed that the actual temperature in the data centre has not changed.
Looking at the output below, it appears to me that the alarm is false, as the FPC 1 CPU temperature is only 39°C, way below the yellow warning theshold of 80°C:
> show system alarms 1 alarms currently active Alarm time Class Description 2016-03-29 12:13:16 CEST Minor FPC 1 CPU Temp Sensor Too Warm> show chassis environment Class Item Status Measurement Power FPC 0 Power Supply 0 OK FPC 1 Power Supply 0 OK Temp FPC 0 CPU OK 37 degrees C / 98 degrees F FPC 0 Exhaust Area OK 55 degrees C / 131 degrees F FPC 0 EX-PFE1 OK 37 degrees C / 98 degrees F FPC 0 EX-PFE2 OK 38 degrees C / 100 degrees F FPC 0 Local Intake OK 31 degrees C / 87 degrees F FPC 0 Remote Intake OK 35 degrees C / 95 degrees F FPC 0 GEPHY1 OK 40 degrees C / 104 degrees F FPC 0 GEPHY2 OK 45 degrees C / 113 degrees F FPC 0 GEPHY3 OK 55 degrees C / 131 degrees F FPC 0 GEPHY4 OK 60 degrees C / 140 degrees F FPC 0 GEPHY5 OK 60 degrees C / 140 degrees F FPC 0 GEPHY6 OK 70 degrees C / 158 degrees F FPC 0 GEPHY7 OK 70 degrees C / 158 degrees F FPC 0 GEPHY8 OK 65 degrees C / 149 degrees F FPC 0 GEPHY9 OK 70 degrees C / 158 degrees F FPC 0 GEPHY10 OK 60 degrees C / 140 degrees F FPC 0 GEPHY11 OK 50 degrees C / 122 degrees F FPC 0 GEPHY12 OK 55 degrees C / 131 degrees F FPC 1 CPU OK 39 degrees C / 102 degrees F FPC 1 Exhaust Area OK 60 degrees C / 140 degrees F FPC 1 EX-PFE1 OK 39 degrees C / 102 degrees F FPC 1 EX-PFE2 OK 44 degrees C / 111 degrees F FPC 1 Local Intake OK 31 degrees C / 87 degrees F FPC 1 Remote Intake OK 39 degrees C / 102 degrees F FPC 1 GEPHY1 OK 45 degrees C / 113 degrees F FPC 1 GEPHY2 OK 50 degrees C / 122 degrees F FPC 1 GEPHY3 OK 50 degrees C / 122 degrees F FPC 1 GEPHY4 OK 65 degrees C / 149 degrees F FPC 1 GEPHY5 OK 65 degrees C / 149 degrees F FPC 1 GEPHY6 OK 70 degrees C / 158 degrees F FPC 1 GEPHY7 OK 65 degrees C / 149 degrees F FPC 1 GEPHY8 OK 75 degrees C / 167 degrees F FPC 1 GEPHY9 OK 70 degrees C / 158 degrees F FPC 1 GEPHY10 OK 55 degrees C / 131 degrees F FPC 1 GEPHY11 OK 55 degrees C / 131 degrees F FPC 1 GEPHY12 OK 55 degrees C / 131 degrees F Fans FPC 0 Fan 1 OK Spinning at normal speed FPC 0 Fan 2 OK Spinning at normal speed FPC 1 Fan 1 OK Spinning at normal speed FPC 1 Fan 2 OK Spinning at normal speed> show chassis temperature-thresholds Fan speed Yellow alarm Red alarm Fire Shutdown (degrees C) (degrees C) (degrees C) (degrees C) Item Normal High Normal Bad fan Normal Bad fan Normal FPC 0 CPU 60 70 60 50 70 60 FPC 0 Exhaust Area 60 70 60 50 70 60 FPC 0 EX-PFE1 30 45 60 50 70 60 FPC 0 EX-PFE2 60 70 60 50 70 60 FPC 0 Local Intake 60 70 60 50 70 60 FPC 0 Remote Intake 60 70 60 50 70 60 FPC 0 GEPHY1 60 70 60 50 70 60 FPC 0 GEPHY2 60 70 60 50 70 60 FPC 0 GEPHY3 60 70 60 50 70 60 FPC 0 GEPHY4 60 70 60 50 70 60 FPC 0 GEPHY5 60 70 60 50 70 60 FPC 0 GEPHY6 60 70 60 50 70 60 FPC 0 GEPHY7 60 70 60 50 70 60 FPC 0 GEPHY8 60 70 60 50 70 60 FPC 0 GEPHY9 60 70 60 50 70 60 FPC 0 GEPHY10 60 70 60 50 70 60 FPC 0 GEPHY11 60 70 60 50 70 60 FPC 0 GEPHY12 60 70 60 50 70 60 FPC 1 CPU 60 70 80 70 95 85 FPC 1 Exhaust Area 60 70 80 70 95 85 FPC 1 EX-PFE1 60 70 60 50 70 60 FPC 1 EX-PFE2 60 70 60 50 70 60 FPC 1 Local Intake 30 45 60 50 70 60 FPC 1 Remote Intake 60 70 60 50 70 60 FPC 1 GEPHY1 60 70 60 50 70 60 FPC 1 GEPHY2 60 70 60 50 70 60 FPC 1 GEPHY3 60 70 60 50 70 60 FPC 1 GEPHY4 60 70 60 50 70 60 FPC 1 GEPHY5 60 70 60 50 70 60 FPC 1 GEPHY6 60 70 60 50 70 60 FPC 1 GEPHY7 60 70 60 50 70 60 FPC 1 GEPHY8 60 70 60 50 70 60 FPC 1 GEPHY9 60 70 60 50 70 60 FPC 1 GEPHY10 60 70 60 50 70 60 FPC 1 GEPHY11 60 70 60 50 70 60 FPC 1 GEPHY12 60 70 60 50 70 60
Another thing that to me looks rather strange, is that the four following sensors have different thresholds on FPC 0 and FPC 1: EX-PFE1, CPU, Exhaust Area, and Local Intake. Is there any particular reason for that?
Even more suspicious is the fact that the FPC 1 Local Intake thresholds are identical with FPC 0 EX-PFE1. The same is true for FPC 1 CPU and FPC 1 Exhaust Area.
The alarm comes and goes, so I wrote a script to that compares the output from show chassis environment before and after the alarm has changed state. I determined that what triggers the alarm is the FPC 1 Exhaust Area sensor, which fluctuates between 59°C and 60°C. This causes the alarm to clear and activate, respectively. However, that does not make any sense either, as the yellow warning threshold for that particular sensor is 80°C.
Is this a JUNOS 12.3 bug, and if so is it a known one? (I searched, but was unable to find any relevant PR or KB articles.) If not, does some other logical explanation why this alarms goes active exist?