Cisco Support Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Announcements

Welcome to Cisco Support Community. We would love to have your feedback.

For an introduction to the new site, click here. If you'd prefer to explore, try our test area to get started. And see here for current known issues.

New Member

ASR 9010 Temperature Threshold

Hi 

I would like to find the temperature threshold value of ASR9010.

admin show environment temperature, it shows only current temp value.

admin show environment table will only have critical value, no minor and major value.

Could anyone provide me this value?

 

 

Thanks,

 

Tom

 

 

Everyone's tags (1)
1 ACCEPTED SOLUTION

Accepted Solutions
Cisco Employee

Hi Tom,

Hi Tom,

great question! And I think we could have documented this a bit better, so here we go!

The asr9000 hardware measures the temperature at 2 points, at the inlet (like the bottom of the card if vertical) and the "hotspot" which is close to the CPU which is top of the card.

The delta between inlet and hotspot effectively gives a measure of the total heat dissipation of that card and depending on how active it is/packets it forwards (no joke) it would be hotter or dissipating more heat.

The environmental logic of the shelf manager has 2 thresholds it would look at:

! #define RSP_STT 93  // RSP STT 

! #define LC_STT  90  // LC STT

These define the shutdown temperature thresholds when the hotspot exceeds these temperatures.

These values are not configurable and fixed in the SW.

Many people would like to have a yellow alarm to get an identification when the temperature reaches a certain (configurable?) threshold so that we can start looking into this. That is something in the works, but not there yet.

The STT (shutdown temperature thresholds) are RED and always be unconfigurable otherwise you will create a toaster oven out of your 9k :)

So what does inlet mean and what temperature should it be at? Another common question:

Ideally you want the temperature as low as possible, but that puts strain on your AC and env costs. If it is too high, but ensuring it is below the STT at the hotspot, the power consumption of the device goes up because more heat means more amps and also higher fan speeds, which also cost power.

So it is a trade off between AC cost and power draw of your device.

Also dirty filters or inconvenient airflow of the cabinets can affect the inlet temperature.

Then the airdensity matters also. If I have a medium inlet temp, but low air density, obviously I can't cool enough so my hotspot could be higher then a device on a location with a higher air density.

In short, there is no minimum or recommended inlet value as long as it can provide enough cooling to the card. The cooler the better, obviously, but trade of cost between AC and power draw and dependent on altitude/air density also.

How to go about it in the absence of the YELLOW alarm? You could poll the entity sensor mib for the hotspot temps and mark a 10% (or whatever range you prefer) and out of out mgmt station generate an alert when the hotspot temp is x% below the STT values provided.

Hopefully this gives some guidance.

regards

xander

~~~

Xander Thuijs CCIE #6775

Principal Engineer ASR9000/XR

Xander Thuijs CCIE #6775 Principal Engineer ASR9000, CRS, NCS6000 & IOS-XR
14 REPLIES
Cisco Employee

Hi Tom,

Hi Tom,

great question! And I think we could have documented this a bit better, so here we go!

The asr9000 hardware measures the temperature at 2 points, at the inlet (like the bottom of the card if vertical) and the "hotspot" which is close to the CPU which is top of the card.

The delta between inlet and hotspot effectively gives a measure of the total heat dissipation of that card and depending on how active it is/packets it forwards (no joke) it would be hotter or dissipating more heat.

The environmental logic of the shelf manager has 2 thresholds it would look at:

! #define RSP_STT 93  // RSP STT 

! #define LC_STT  90  // LC STT

These define the shutdown temperature thresholds when the hotspot exceeds these temperatures.

These values are not configurable and fixed in the SW.

Many people would like to have a yellow alarm to get an identification when the temperature reaches a certain (configurable?) threshold so that we can start looking into this. That is something in the works, but not there yet.

The STT (shutdown temperature thresholds) are RED and always be unconfigurable otherwise you will create a toaster oven out of your 9k :)

So what does inlet mean and what temperature should it be at? Another common question:

Ideally you want the temperature as low as possible, but that puts strain on your AC and env costs. If it is too high, but ensuring it is below the STT at the hotspot, the power consumption of the device goes up because more heat means more amps and also higher fan speeds, which also cost power.

So it is a trade off between AC cost and power draw of your device.

Also dirty filters or inconvenient airflow of the cabinets can affect the inlet temperature.

Then the airdensity matters also. If I have a medium inlet temp, but low air density, obviously I can't cool enough so my hotspot could be higher then a device on a location with a higher air density.

In short, there is no minimum or recommended inlet value as long as it can provide enough cooling to the card. The cooler the better, obviously, but trade of cost between AC and power draw and dependent on altitude/air density also.

How to go about it in the absence of the YELLOW alarm? You could poll the entity sensor mib for the hotspot temps and mark a 10% (or whatever range you prefer) and out of out mgmt station generate an alert when the hotspot temp is x% below the STT values provided.

Hopefully this gives some guidance.

regards

xander

~~~

Xander Thuijs CCIE #6775

Principal Engineer ASR9000/XR

Xander Thuijs CCIE #6775 Principal Engineer ASR9000, CRS, NCS6000 & IOS-XR
New Member

Hi Xanderthanks for your

Hi Xander

thanks for your prompt reply

 

Tom

Does it have any

Does it have any recommendation since what is value I should put more air conditioner?

PlaWanSai RMUTT CPE IX
Cisco Employee

Check table A-2http://www

Check table A-2

http://www.cisco.com/c/en/us/td/docs/routers/asr9000/hardware/installation/guide/asr9kHIGbk/asr9kIGapaspecs.html

 

We recommend staying between 5 and 40C, this can be checked with admin show environment temperature.

 

Thanks,

Sam

New Member

Can you advise where to find

Can you advise where to find the table? The document you posted seems to reference it, but I can not find a table A-2 anywhere...

Thanks

Cisco Employee

evan here you go:

evan here you go:

Table A-2 Environmental Specifications

Description
Value

Operating Temperature (Nominal):

41° to 104°F
(5° to 40°C)

Operating Temperature
(Short term)2:

23° to 131° F
(–5° to 55°C)

Humidity

Operating: 10 to 85 percent noncondensing

Nonoperating: 5 to 95 percent noncondensing

Altitude

Operating: 0 to 13,000 ft (0 to 4,000 m)

Nonoperating: 0 to 15,000 ft (0 to 4,570 m)

Power Dissipation

750 W maximum

Acoustic noise

70 dB at 80.6°F (27°C) maximum

Shock

Operating (halfsine): 21 in/sec (0.53 m/sec)

Nonoperating (trapezoidal pulse): 20 G3, 52 in/sec (1.32 m/sec)

Vibration

Operating: 0.35 Grms4 from 3 to 500 Hz

Nonoperating: 1.0 Grms from 3 to 500 Hz

2.Short-term refers to a period of not more than 96 consecutive hours, and a total of no more than 15 days in a year. (This refers to a total of 360 hours in any given year, but no more than 15 occurrences during that 1-year period.).

3.G is a value of acceleration, where 1G equals 32.17 ft/sec2 (9.81 m/sec2).

4.Grms is the root mean square value of acceleration.

Xander Thuijs CCIE #6775 Principal Engineer ASR9000, CRS, NCS6000 & IOS-XR
New Member

Hi Xander and ASR9K Expert,

Hi Xander and ASR9K Expert,

For RSP or RP, what sensor do we have to monitoring?

Is it fine to monitoring only for 'Inlet' sensor and set the threshold not over 40C

If the both sensors must to be montiored, how about the threshold of 'Hostspot' on RSP/RP ? or Can i use the config in SW (81 degreeC) as threshold?

Thanks in advance

Cisco Employee

hi rojer, it is the hotspot

hi rojer, it is the hotspot that is most important/critical.

if the hotspot goes over the set threshold, the card will eventually shutdown.

regards

xander

Xander Thuijs CCIE #6775 Principal Engineer ASR9000, CRS, NCS6000 & IOS-XR
New Member

Hi Xander,

Hi Xander,

Thanks for reply. So the hotspot sensor should also be monitored and set threshold < 80 degreeC correct?

Cisco Employee

yeah the inlet is merely to

yeah the inlet is merely to show you the intake temperature, it can be warm, if the devices are not hot at all or when the air is dense. the lower the temp is the more heat it can absorb obviously.

the hotspot is definitely more critical. would set a marker on 80 for yellow and 90 for red.

XR6 added a nice functionality to manually configure temperature thresholds for yellow and red via CSCun79821

cheers!

xander

Xander Thuijs CCIE #6775 Principal Engineer ASR9000, CRS, NCS6000 & IOS-XR
New Member

Thanks Xander

Thanks Xander

New Member

Hi Alexander,

Hi Alexander,

We found this log on our ASR9k, not sure if the room temperature is not cool enough because we see on RSP its quite normal, we suspecting the linecard carrying lot of traffic.

LC/0/0/CPU0:2016 Apr 18 02:54:07.165 WIB: pfm_node_lc[299]: %PLATFORM-TEMPERATURE_ALARM-3-DIE_OVERTEMP_MAJOR : Set|envmon_lc[172120]|0x103c005|Alarm:Major Sensor:FIA5 Location:0/0/CPU0
LC/0/0/CPU0:2016 Apr 18 02:56:21.632 WIB: pfm_node_lc[299]: %PLATFORM-TEMPERATURE_ALARM-3-DIE_OVERTEMP_MAJOR : Clear|envmon_lc[172120]|0x103c005|Alarm:Major Sensor:FIA5 Location:0/0/CPU0
RP/0/RSP1/CPU0:2016 Apr 18 02:56:22.781 WIB: envmon[208]: %PLATFORM-ENVMON-4-DIE_OVERTEMP_MAJOR : Slot 0/0/CPU0 sensor FIA5 temperature 103.0C cleared threshold 108.0C

RP/0/RSP1/CPU0:PE01-NGN-JKT-MNCPDC#admin show environment temperatures
Mon Apr 18 11:02:05.634 WIB
R/S/I   Modules Sensor                  (deg C)
0/0/*
        host    die_NP0                 81.5
        host    die_NP1                 66.5
        host    die_NP2                 60.0
        host    die_NP3                 58.5
        host    die_NP4                 56.5
        host    die_NP5                 56.5
        host    die_FIA0                71.0
        host    die_FIA1                49.5
        host    die_FIA2                59.5
        host    die_FIA3                90.0
        host    die_FIA4                61.0
        host    die_FIA5                105.0
        host    die_FabSwitch           89.0
        host    db_np_0                 71.6
        host    db_np_1                 55.5
        host    db_skt_0                53.9
        host    db_inlet                33.1
        host    bb_inlet                32.8
        host    bb_hotspot_0            51.8
        host    bb_exhoust_0            59.9
        host    bb_exhoust_1            58.3
        host    bb_hotspot_1            67.7
        host    Inlet0                  32.5
        host    Hotspot0                50.6
                              



0/RSP0/*
        host    die_CPU                 32.0
        host    die_FabArbiter0         37.0
        host    die_FabArbiter1         39.0
        host    die_FIA                 41.0
        host    die_FabSwitch0          42.5
        host    die_FabSwitch1          42.5
        host    temp_u275               23.0
        host    temp_u276               32.7
        host    temp_u278               33.0
        host    temp_u277               28.6
        host    temp_u279               31.5
        host    temp_u280               42.1
        host    temp_u281               29.6
        host    temp_u282               44.5
        host    Inlet0                  23.2
        host    Hotspot0                32.5
                              
0/RSP1/*
        host    die_CPU                 32.0
        host    die_FabArbiter0         37.0
        host    die_FabArbiter1         39.0
        host    die_FIA                 44.0
        host    die_FabSwitch0          40.0
        host    die_FabSwitch1          40.0
        host    temp_u275               22.8
        host    temp_u276               32.5
        host    temp_u278               31.3
        host    temp_u277               27.5
        host    temp_u279               31.3
        host    temp_u280               41.3
        host    temp_u281               29.4
        host    temp_u282               43.8
        host    Inlet0                  23.0
        host    Hotspot0                32.2
                              



0/7/*
        ep0     Inlet0                  27.6
        ep0     Hotspot0                36.1
                              
        host    die_NP0                 50.0
        host    die_NP1                 54.0
        host    die_FIA0                52.0
        host    die_FIA1                57.0
        host    die_FabSwitch           52.0
        host    Inlet0                  26.1
        host    Hotspot0                42.1
                              
0/FT0/*
        host    Inlet0                  24.6
        host    Hotspot0                29.0
                              
0/FT1/*
        host    Inlet0                  22.8
        host    Hotspot0                26.4

Also from "admin show environment table", we see there is major and critical Hi value, In what condition that linecard will shutdown/reload ? Please clarify and give some comment.

Thanks.

Cisco Employee

sounds like you have a 36x10

sounds like you have a 36x10 in slot 0 in a 9010.

the FIA number 5 is running way hotter then all the other components. I see the inlet is rather high for this slot also compared to the other slots.

Can you check the fan speeds and see if there is adequate airflow? I think there will be considering the inlet vs hotspot/temps on the RP's.

I would check the airfilter to make sure they are able to provide the right airflow.

If there are any high powered optics inserted in the card, we may need to move the card to a more centered slot for it to have it benefit the max airflow, which is slightly higher in slots 3 and 4 around the RSP's.

xander

Xander Thuijs CCIE #6775 Principal Engineer ASR9000, CRS, NCS6000 & IOS-XR
Cisco Employee

Line card will be shutdown 24

Line card will be shutdown 24 hours after any of the sensors hits critical temperature threshold. Critical threshold values can be checked through "admin show environment table" command output.

When major threshold is hit, fan's will be automatically run faster to cool down the line card. this is why you see major alarm getting cleared.

Thanks,

-Phadke

1711
Views
20
Helpful
14
Replies