Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I've got a UTP Cat-6 ethernet installation with a Linux server in one end, my PC in the other and three gigabit switches in between. Recently I changed the cable from a normal Cat-5 to an UTP Cat-6, but it doesn't seem to be working properly. All the switches I have lit another led or change the led color if the conection is either 10/100 or gigabit, but none of the switches recognice the new cable as gigabit (they lit as it were a 10/100).

The length of the cable between switches is no more than 20m and I've tried with different Cat-6 connectors (shielded and unshielded).

I did some tests yesterday and this is the result I got:

E:\Downloads>iperf -c 192.168.1.111 -d
------------------------------------------------------------
Server listening on TCP port 5001
TCP window size: 8.00 KByte (default)
------------------------------------------------------------
------------------------------------------------------------
Client connecting to 192.168.1.111, TCP port 5001
TCP window size: 63.0 KByte (default)
------------------------------------------------------------
[188] local 192.168.1.204 port 49699 connected with 192.168.1.111 port 5001
[ ID] Interval Transfer Bandwidth
[188] 0.0-10.0 sec 113 MBytes 94.6 Mbits/sec
[144] local 192.168.1.204 port 5001 connected with 192.168.1.111 port 47400
[ ID] Interval Transfer Bandwidth
[144] 0.0-10.0 sec 113 MBytes 94.1 Mbits/sec

The UTP cable I bought has everywhere (in the box and in the cable itself) that its a Cat-6 UTP cable, so it should perform much better than the above example. As for the switches, I've test them with a premade gigabit cable and they lit as it was a gigabit connection.

Any idea where could be the problem? Could the cable be faulty?

EDIT: I've just tested with a LAN tester (more or less like this one) and the cable seems to be ok. All the links light in order and when they're supposed to (both ends are the same).

share|improve this question
    
If you have tested your equipment with a confirmed CAT 6 cable, then my guess is that someone either sold you a bad cable, or a a mislabeled cable. –  Dudemcman Apr 22 '13 at 13:36
    
Did you wire the ends of the cable yourself? If so, what pin-to-pair mapping did you use? –  David Schwartz Apr 22 '13 at 13:41
    
It's just very likely that you have not crimped on the connector correctly, e.g. one or more wires are not connected correctly or the order is wrong on one end. You should test all 8 wires at the connector contacts, e.g. with a simple continuity tester or an ohmmeter. –  Stefan Seidel Apr 22 '13 at 14:08
    
True, I have a tester but I didn't test all the wires, I will run that test and see if its correctly wired or not. –  Peter Apr 22 '13 at 14:56

1 Answer 1

up vote 1 down vote accepted

UTP = Unshielded Twisted Pair, right?

It is possible that you are getting alien crosstalk, i.e. interference from EM spectrum emitted by other devices externally to the cable. In my experience, video cables (VGA, DVI, HDMI) like to do this if you lay them on top of an unshielded ethernet cable.

I used to have problems like this, but then I learned that Category 7 cable isn't incredibly expensive anymore, and is quite well shielded to allow for strong 10 Gigabit signals and can easily support Gigabit even with a lot of external signal attempting to interfere.

Per the tables here, the first Category of ethernet structured cabling that requires shielding per the standard is Cat 6a, but if you can afford Cat 7, I'd try that instead. Buy from a reputable vendor to avoid getting jipped by receiving ineffective shielding or no shielding at all. Since the cable looks the same, an amoral vendor could easily label a standard Cat5 cable as Cat7, when they are extremely different if you were to cut the cable open and look inside.

Cat 7 is also known as Class F so you may have better searching results if shopping for "Class F" vs. "Cat 7".

Lastly, to eliminate the chance of the cable being incorrectly terminated by you, I would recommend that you buy cables that come with the connector professionally manufactured into the cable. This may come at additional cost, but you can certainly buy cable in any length needed for your project pre-terminated, unless you need lengths that exceed reasonable operating length as specified by the standard.

Sources:

Wiki article on Class F
This site, which is a source linked to by Wiki
Wiki article on UTP

P.S. -- A model year 2012 Netgear router rated at gigabit speeds recently purchased by myself, a Netgear N900 / WNDR4500, provided, in the box, at an American retail store, a Cat 7 / Class F cable for connecting the user's modem / gateway to the router. If a consumer-oriented device at gigabit line speed is coming with Cat 7 included in the box, you can imagine how strongly the vendor feels that giving them a shielded cable will result in reduced support calls due to poor performance induced by crosstalk. Thus you are most likely a victim of this very phenomenon.

share|improve this answer
    
As stated in the answer, UTP is s**t (or at least the cable I bought). I've just done the same test with an STP CAT6 cable and got 672 Mbits/sec (84 MBytes/sec) with the same cable... ^^ –  Peter Apr 23 '13 at 17:30
    
So STP fixed it? Cool! –  allquixotic Apr 23 '13 at 18:54

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.