Problem Description

The customer found that several SSR1PEFS8 veneer boards in slot 9 of OSN1500 sites have been reporting TEMP_OVER alarms. The other boards are normal, and the ambient temperature in the server room is around 25 degrees Celsius, which is also normal. the PEFS8 board version is 1.52, and the host software is 5.36.31.15.

Alarm message

TEMP_OVER MJ start 2011-10-09 15:10:50 None NSA NEW_BOARD board=9;01 ;;

Processing

The SSR1PEFS8 board temperature alarm threshold is set low, causing the SSR1PEFS8 board to report TEMP_OVER frequently.

Reset the SSR1PEFS8 single board temperature threshold:

[Setting method]

set boardid 9

//Set the threshold

:cfg-set-bdtempthboardid,750,0

//Query whether the threshold is set successfully

:cfg-get-bdtempthboardid

Print the following message to indicate that the setting is successful

BOARD-TEMP-TH

TEMP-HIGH-TH TEMP-LOW-TH

750 0

Total records :1

Root cause

According to the data returned from the first-line acquisition, it is found that the current temperature of SSR1PEFS8 board is 66.5 degrees Celsius. In addition, the alarm threshold is set to 65 degrees Celsius. Then it can be seen that reporting the TEMP_OVER alarm is normal.

:cfg-get-bdtemp:9

BOARD-TEMP

BID TEMP-NOW

9 665

Total records :1

:cfg-get-bdtemp:9

BOARD-TEMP-TH

TEMP-HIGH-TH TEMP-LOW-TH

650 0

Total records : 1

Meanwhile, according to the current temperature data returned from other single boards, it is found that the current temperature of other single boards is more than 10 degrees Celsius lower than that of SSR1PEFS8. Therefore, it is suspected that the temperature alarm threshold of SSR1PEFS8 single board is set reasonably.

:cfg-get-bdtemp:6

BOARD-TEMP

BID TEMP-NOW

6 475

:cfg-get-bdtemp:7

BOARD-TEMP

BID TEMP-NOW

7 540

Total records :1

To determine the reasonableness of the SSR1PEFS8 temperature alarm threshold in a lab setting: Set up an OSN1500 fully configured network element at 55 degrees Celsius ambient temperature, and test the SSR1PEFS8 key chip and power supply case temperatures with a thermometer:

Power supply temperature: 72.6 degrees

EP3C80 chip temperature: 71.7 degrees

And use navigator to query the current temperature result of SSR1PEFS8 single board:

#9-555:szhw [][][2010-07-24 17:57:25+08:00]>

>>>:cfg-get-bdtemp:9

BOARD-TEMP

BID TEMP-NOW

9 765

Total records :1

Confirm that the board temperature exceeds the temperature threshold and the threshold is not set properly.

Recommendations and Summary

Due to the high power consumption of the PEFS8 board and its distance from the fan, the operating temperature of this board is high. When the factory set the temperature threshold, it was not taken into consideration, so it was still designed according to the default value, which led to the alarm report.