UCS CPU temperature thresholds

[John Sutch]

We have a number of B200 M1 blade servers and we are occasionally seeing 'temperature: upper-non-critical' for the CPUs on these blades. This seems to co-incide with us running benchmarking software against the blades so I am not too concerned, but I would like to know what the threshold settings are. We don't have any threshold policies configured, so I presume that there are a number of defaults. However, I haven't been able to determine what these defaults are. Can anyone help?

Thanks,     John.

[abbharga · Accepted solution]

Hi John,

For each componenet there are fixed limits which cannot be modified.

There is a quick and easy way for you to check these values:

log onto to the CLI:

#connect cimc x/y

(x - chassis / y - blade #)

#sensors

Look for the values:

P2_TEMP_SENS   

P1_TEMP_SENS

The last three column values will give you the threshold.

Here is the sample for a B200 M1 blade:

-A# connect cimc 1/1

Trying 127.5.1.1...

Connected to 127.5.1.1.

Escape character is '^]'.

CIMC Debug Firmware Utility Shell [ support ]

[ help ]# sensors

Sensor Name      | Reading | Unit         | Status | LNR     | LC      | LNC     | UNC     | UC      | UNR     |

=================|=========|==============|========|=========|=========|=========|=========|=========|=========|

P2_TEMP_SENS     | 47.000  | degrees C    | OK     | na      | na      | na      | 80.000  | 85.000  | 90.000  |

P1_TEMP_SENS     | 47.000  | degrees C    | OK     | na      | na      | na      | 80.000  | 85.000  | 90.000  |

LEGEND:

LNR: Lower Non Recoverable Threshold

LC : Lower Critical Threshold

LNC: Lower Non Critical Threshold

UNC: Upper Non Critical Threshold

UC : Upper Critical Threshold

UNR: Upper Non Recoverable Threshold

Hope this helps!

./Abhinav

[John Sutch]

Thanks Abhinav. That was a very quick response!! One more question: if the temperature exceeds the upper crtical threshold, does the server shut down?

[abbharga]

John,

In general there is no automatic shutdown of the server or the OS. In case of temprature crossing each limit the fans adjust and spin higher to coll the CPU or the blade down.

The servers keep going untill the CPUs reach their down protection temprature (which varies from CPU to CPU). Even before that the CPUs start slowing down to reduce the temprature.

In very adverse conditions (very high tempratures) the power supplies start shutting down and then shutting the blades (which is very rare).

./Abhinav

[John Sutch]

Thanks for the comprehensive answer Abhinav.