We have recently done a change in our network by using CAT 6 cabling. Now I would like to measure the speed difference when I use Cat 5 or Cat5 e and Cat 6. Is there a way to find out the same?
Asked
Active
Viewed 418 times
5 Answers
2
CAT5e is perfectly capable of supporting 1GB connectivity. Most do not realize that the issue is typically not in the transmission media, rather at the points where IO creates bottlenecks.
slashdot
- 651
- 5
- 7
0
If you have a file server that you know is not the bottle neck, simply transfer a big compressed file a few times and time it while there is not much other network activity.
In some cases Cat5 can work up to a gigabit, and if that was your case, you will probably not notice a huge difference.
Gepeto
- 236
- 1
- 5
0
Are you running gigabit switches/network cards? If not, your not going to notice any difference. If you are, then depending on disc i/o, you should be able to transfer files quicker- test from server > server.
AliGibbs
- 2,303
- 20
- 34
-
Yes we are running gigabit switches. How do I measure the disc i/o to transfer the file from one server to another. I have one linux server and the other is a windows server. – user45774 Feb 22 '11 at 13:41
0
You can use 'netcat' on Linux to test pure network speed, without having to worry about approximations using file transfers (limited by disk IO).
Craig Mason
- 133
- 1
- 6
-
In this scenario, if you had Linux, would you use /dev/zero on the sender end and /dev/null on the receiving end? (or just nul on the recevier if it runs Windows) – jftuga Feb 18 '11 at 13:48
-
1Correct. However, [iperf](http://iperf.sourceforge.net/) is actually a better solution nowadays. (I forgot about this when replying!) – Craig Mason Feb 21 '11 at 10:54
0
The most limiting factor won't be the cabling - but disc-performance and the servers/clients and their configuration (+ switches/routers) matter.
The biggest impact on the "faster" cables will be fewer errors - at the same speed...
tsg
tsg
- 284
- 1
- 1