Led Alarms on A9K-RSP-4G how to tshoot

[g3doubt]

Hello,

Im new on IOS-XR and ASR9K platforms so I need help to understand how to tshoot the led alarms on RSP (for this case critical and Major) running version 5.3.4. I tried to extract information using show alarms brief and admin show alarms but it didn´t support it.

I´ve checked the environment of the ASR9K and seems normal (using show environment all)

My questions are the following:

• The RSP led alarms indicates a logical problem or an evironment problem?
• A log could trigger those led alamrs?
• The Warning on blocked porcesses can trigger the led alarm?
• Which commnds I can use to have more information about those led alarms.

Thanks and regards

[smilstea]

Show alarms helps but show pfm loc all well give all alarms and their severity per slot.

Sam

[g3doubt]

@smilstea Thanks, that helped me a lot.

Most of the alarm are low RX power on SFP and XFP but one is for a failure on a diagnostic of a LC:---------------------
CURRENT TIME: 
PFM TOTAL: 2 EMERGENCY/ALERT(E/A): 1 CRITICAL(CR): 1 ERROR(ER): 0
-------------------------------------------------------------------------------
Raised Time |S#|Fault Name |Sev|Proc_ID|Dev/Path Name |Handle
--------------------+--+-----------------+---+-------+--------------+----------
 2022|1 |LC_NP_LOOPBACK_FA|E/A|159830 |Line card NPU |0x2000006

CPU0:

Overall diagnostic result: MINOR ERROR
Diagnostic level at card bootup: bypass

Test results: (. = Pass, F = Fail, U = Untested)

1 ) PHYCtrlScratchRegister ----------> .
2 ) PortCtrlScratchRegister ---------> .
3 ) CPUCtrlScratchRegister ----------> .
4 ) NPScratchRegister ---------------> .
5 ) BridgeScratchRegister -----------> .
6 ) FIAScratchRegister --------------> .
7 ) EccSbeTest ----------------------> U
8 ) LcEobcHeartbeat -----------------> .
9 ) NPULoopback ---------------------> .
10 ) NPULoopbackUseRxPath ------------> F
11 ) FPDimageVerify ------------------> U

Now my question is the following: How do I solve this failure of the diagnostic? Do I need to reset (hw or SW) the LC?

I´ve read that it is needed to reload the entire chassis

Regards

[smilstea]

That error indicates there is an error on the punt inject path on the lc itself to its local cpu when testing that path. You can try reloading the lc, but really we need to see what else is going on in this box as well as what type of card it is and the patches for 5.3.4 you have applied. We have fixed a number of issues in 5.3.4 and later versions related to issues like this, also like I mentioned it could be transient and a reload will correct it.. In either case 5.3.4 is a rather old software version that isn't supported anymore, and depending on your hardware the hw possibly isn't either. I would try a reload of the LC, if that doesn't help we can look into applying the latest SMUs/SP for 5.3.4 or a SW upgrade if possible with your hardware.

Sam

[aemmanue]

• The RSP led alarms indicates a logical problem or an evironment problem?
  • Can be both example Fan tray / power failure can trigger as well 
• A log could trigger those led alamrs?
  • Usually yes 
• The Warning on blocked porcesses can trigger the led alarm?
  • No unless it affects some standby ready state. 

you can try it will give the alarm reasons as well : 

show environment leds details

!

check the following as well : 
show platform 
show redundancy

Regards,

Gobinath. 

[Jose Nava Nieto]

Hello Gerardo, You can find the RSP LED Definitions on the link below.
It provides a detailed description of the LED alarms on the RSPs and other modules

https://www.cisco.com/c/en/us/td/docs/iosxr/asr9000/hardware-install/overview-reference/b-asr9k-overview-reference-guide/b-asr9k-overview-reference-guide_chapter_01.html#con_819068.