Strange Results TDR Cable-Test

[jdparsons]

Has anyone seen all remote pairs showing as "Pair A" and Not Supported? Below are the results from the switch. For reference this is on a Catalyst 3850 with an Aironet 1852i access point on the distant end. This AP has been going down every few days for several hours at a time.

TDR test last run on: January 23 09:13:16

Interface.   Speed   Local pair           Pair length        Remote pair            Pair status
--------- ----- ---------- ------------------ ----------- --------------------
Gi1/0/10 1000M

                                 Pair A                  N/A                Pair A                 Not Supported
                                 Pair B                  N/A                Pair A                 Not Supported
                                 Pair C                  N/A                Pair A                 Not Supported
                                 Pair D                  N/A                Pair A                 Not Supported

#show power inline gi 1/0/10 de
Interface: Gi1/0/10
Inline Power Mode: auto
Operational status: on
Device Detected: yes
Device Type: cisco AIR-AP1852I-B
IEEE Class: 4
Discovery mechanism used/configured: Ieee and Cisco
Police: off

Power Allocated
Admin Value: 30.0
Power drawn from the source: 24.8
Power available to the device: 24.8

Actual consumption
Measured at the port: 5.3
Maximum Power drawn by the device since powered on: 10.2

[balaji.bandi]

That is strange,  is this going via any patch pannel ?

@leo wrote nice document :

https://community.cisco.com/t5/networking-knowledge-base/how-to-use-time-domain-reflectometer-tdr/ta-p/3119327

[jdparsons]

Yes it is going through a total of 2 patch panels as are several other APs on the same switch, none have the above result. What makes this more odd is you would think it'd be a cabling issue, but the AP is currently up and showing this. I'm stumped.

I've referenced the above document many times. It is indeed very helpful.

[balaji.bandi]

i have seen when they going via patch pannel not got as expected.

[Leo Laohoo]

---

@jdparsons wrote:
For reference this is on a Catalyst 3850 

---

Any switch running on IOS-XE, regardless the firmware version, TDR is completely and utterly broken.  Any result is no longer accurate nor trustworthy.  

[jdparsons]

I've noticed a lot like that after doing a site visit and testing the cables with a Fluke. Just never seen one like this.

[Leo Laohoo · Accepted solution]

Here are some examples of some "broken" state: (Above)  Pair "A" distance is (significantly) more than Pairs B, C and D.

(Above)  Pair A & B distance is "0".

9300, IOS-XE version 17.9.4a

To anyone reading this response:  If you are in NAM/LATAM/SJ/RTP or Europe/Middle East/Africa time-zone, take screenshots of "bizarre" TDR result and raise a TAC Case.

[Rich R]

As Leo says TDR on IOS-XE is completely unreliable.  We've got a TAC case open for it on 9300 switches which has been running since April last year.  TAC now agree that the results can't be trusted.  The BU have a case open with their chipset vendor (a few months now) and we're waiting for them to come back to us about it.  Interesting fact - we've found that if there's only 1 cable connected to the switch TDR will usually work well.  As soon as you connect other cables the results (even on the port which was working by itself) go random.  And for those mentioning patching - it happens even with a direct cable (no patching) and with multiple different kinds of devices connected (routers, UCS servers, laptops) - we've been through ALL those tests with TAC multiple times (because at first they couldn't believe how bad their own kit was).  Most of the TAC engineers have evidently never tested  this themselves on a switch so they didn't know it didn't work.  One of the TAC engineers admitted he was embarrassed that he'd been telling customers their cables were faulty when all this time it was their TDR test not working.

[eglinsky2012]

Rich, do you have a bug ID you can share? I'd like to keep updated on that issue. And what code do you run?

I haven't noticed any strange behaviors with a TDR test. I believe a TDR test is only accurate if the device on the other end (the WAP in this case) is unplugged. I don't do this often since we have LinkRunners we usually do TDR tests with, but when I do, it's either indicated the cable was good or reported an actual issue. We were on 16.12.7 until the holidays, now on 17.6.4 but haven't done any TDR tests that I can recall.

[Leo Laohoo]

---

@eglinsky2012 wrote:
And what code do you run?

---

TDR has been broken since 16.3.X (and later).

[Rich R]

It took them a while but they eventually opened a bug: https://bst.cloudapps.cisco.com/bugsearch/bug/CSCwd97177

The description leaves room for improvement and it's not even logged against the correct IOS though it no doubt also affects 17.3.5.  According to Cisco the test should work when devices are plugged in except with certain specific port/device configurations - and it generally does if nothing else is plugged in.

[Leo Laohoo]

Thanks for this, @Rich R.  I might start sending TAC some of our TDR capture details.  

[Leo Laohoo]

---

@Rich R wrote:
CSCwd97177

---

Added a 2nd TDR Bug ID:  CSCvw97924