Cable Diagnostics test

[michael brock]

Hello, 

Our team is having an internal discussion regarding the Cable diagnostics test "test cable-diagnostics tdr interface xxx" command and if POE should be turned off when running the test. One group believes that if the end device requires POE, you leave it on when testing and if its not a POE device, turn it off before testing, then reenabling after test is complete.

Thoughts?

[Flavio Miranda]

Who won?

[michael brock]

Thats why I posted this, to see if it mattered. Who won has yet to be established.

[Leo Laohoo]

Depends on the OS.  If the OS is IOS-XE, depending on the firmware version. 

Typically, TDR in classic IOS is fairly accurate.  In IOS-XE if the firmware is 3.X.X, the TDR result is also accurate.  However, if the IOS-XE version is 16.X.X and/or 17.X.X, the TDR result is not to be trusted because it is no longer accurate.  See examples below:

(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

Because all of our IOS-XE switches are PoE, we found out that turning off PoE (power inline never) first and then launching the TDR test will yield a better result. 

Currently, there are two Bug IDs:  CSCvw97924 & CSCwd97177

Initially, we were told that the bug is due to the Broadcom chip.  From another Cisco staffer, CSCvw97924 & CSCwd97177 will never be fixed and blaming Broadcom is just a way to divert everyone's attention.  Why blame Broadcom when it was working if the switch is on 3.X.X?