We've inherited an existing network that is a mixture of a cat5e, cat6, with switches and rj45 couplers that tie everything together.

Is there a way to test a local network speed from one side of the network say by the modem to the other side of the network at its furthest point ?

  • Before you waste too much time, the cabling works, or it doesn't. You won't find any difference between Category-5E or Category-6 cabling at speeds up to 1 Gb. Remember that a cable system 's rating is the rating of the lowest rated link in the system, so adding Category-6 cables to a Category-5E system is a waste of money. The RJ-45 couplers you mention are completely non-standard and should never be used in network cabling.
    – Ron Maupin
    Oct 3, 2015 at 18:29
  • @RonMaupin when you say "The RJ-45 couplers you mention are completely non-standard.." in principle arnt they the same as a wall outlet data point ? Whats the danger / downside of them ?
    – sam
    Oct 4, 2015 at 23:50
  • Structured cable systems are designed to pass a suite of tests to claim a category rating. The design calls for horizontal cable in a single link up to 90 meters using solid-core cable. On each end, you have a single link patch cord of stranded cable up to 5 meters. Couplers, introduce sections of cable which are not twisted, and extra points of impedance mismatches, none of which are taken into account for the design. You will not pass the cable test suite using couplers in a cable system.
    – Ron Maupin
    Oct 5, 2015 at 0:33
  • You could get away with things like couplers in the old Category-3 , 10BASE-T days, but modern ethernet and cable systems are very sensitive to non-standard cabling due to the very high frequencies and tight timing employed.
    – Ron Maupin
    Oct 5, 2015 at 0:38

2 Answers 2


iPerf or Tamosoft throughput test. As a primarily Mac shop I currently find TTT more convenient, as iPerf builds for Mac are somewhat limited to specific OS versions or "build it yourself". TTT is a straightforward (and cross-platform) download and run.

...but iPerf is by far the more commonly/widely accepted method among network professionals.

Have used both, not affiliated with either. The old "copy a large file from a computer here to a computer there" method is also informative and straightforward.

To the horror of cable salespeople, you will probably not find any noticeable difference between Cat5e and Cat6 segments, unless a segment of either type is kinked or otherwise defective...Or plugged into a non-gigabit switch...


If you use linux or windows, you can test with Iperf, and server-client application that easily can do what you want. You can get information here:


Example: Client: iperf -c ip_Server Server: iperf -s

Also you can change the parameters to do a more strong test

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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