We have quite a few Nortel Ethernet Routing Switch 5500s deployed throughout our organization. There’s a great new benefit in using the new hardware to help us test the cable plant remotely.
Here’s the text from the Nortel manual;
Testing cables with the Time Domain Reflectometer
With Release 5.0 software, the Nortel Ethernet Routing Switch 5500 Series is equipped with a Time Domain Reflectometer (TDR). The TDR provides a diagnostic capability to test connected cables for defects (such as short pin and pin open). You can obtain TDR test results from the CLI or the JDM. The cable diagnostic tests only apply to Ethernet copper ports; fiber ports cannot be tested. You can initiate a test on multiple ports at the same time. When you test a cable with the TDR, if the cable has a 10/100 MB/s link, the link is broken during the test and restored only when the test is complete. Use of the TDR does not affect 1 GB/s links.
Note: The accuracy margin of cable length diagnosis is between three to five meters. Nortel suggests the shortest cable for length information be five meters long.
Unfortunately this feature is ONLY available on the 5510, 5520 and 5530 switches.
Using Device Manager you’ll find the option on the port settings (a tab to the right labeled “TDR”). You can also use the following CLI commands;
tdr test <portlist> show tdr <portlist>
Off topic but… On occastion we’ve had end-users plug a cable back into a 5520, which of course results in a loop. The docs say that you can turn off auto MDI/MDI-X by turning off auto-negotiation. Is there a way to turn off auto MID/MID-X without turning off all auto negotiation?
Michael McNamara says
I was just going over all the comments and realized I never responded to this comment, please accept my apologies whomever you are (were?).
I don’t believe there is a way to disable the MDI/MDI-X feature short of disabling auto-negotiation (which really isn’t an option). In my organization we’ve taken to the practice of enable Spanning Tree on every edge switch port (notice I said “edge”, I say this meaning we exclude running Spanning Tree on any uplinks to the core). We also enable FastStart on every port to avoid the 30 second delay in forwarding traffic before Spanning Tree transitions to the forwarding mode.
Have upgraded our 5510 switches from v4 to v6 – our 8006 switches remains on v3,5 for two more weeks. In total; 45 stacks totaling about 98 switches. All stacks but one went fine, when some users conects to that stack they get no contact. So I ran the TDR test: it reported that of 92 available ports 86 ports had cable problems….seams a “bit” odd. So what I wonder is: how reliable is that tool?
Michael McNamara says
It sounds to me like you might have a problem somewhere. Assuming the cabling was in place and working before the upgrade I wouldn’t think the problem lies with the cabling. Your test results are probably being affected by whatever issue the stack is already experiencing. Have you cold booted the entire stack? It’s possible that you have a hardware issue… is there anything of interest in the log? I’m assuming you loaded software release 6.0.3 and are aware of the VLACP issues with 6.0.0/6.0.1.
Here’s the results from an ERS 5520 with an 1140e connected to port 1/17;