Good Day. I just purchased a Plugable Gigabit USb 3.0 USB3-E1000, which i have installed on a USB 3 port on a Linux Ubuntu 12.04 system. It detected it ok and according to the connection information is 1000 Mb/s.
Whenever i try to copy a file to a Windows or Linux system i just get a transfer speed of 10-13 MB/s.
I have tried this on two different location, which means, 2 different Gigabit routers, 2 different set of Cat 5e+ cables, 2 different systems both with GB nics, and 2 different OS.
I dont know what am i doing wrong. Could anyone point me to the right direction?
Hi Juan - thanks for posting. We’ll try to figure out what might be happening.
10-13 MBs would be roughly what you’d see if there were a Fast Ethernet switch anywhere on the network between the two sides. So that’s one thing to look for.
On a gigabit network, you should see about 5x that speed in practice – but that’s measuring with a low level tool (http://www.thegeekstuff.com/2011/12/l…). Copy to/from a Windows box often shows a lower transfer rate in the Windows UI, just because of the other components involved and how it’s measuring.
Can you say what tool you’re using to measure the 10-13MBs? And have you ever had gigabit speeds between to devices on this network, or might there be a legacy fast ethernet switch left behind somewhere in the chain?
Thanks and best wishes!
Im not using any tool in particular to measure the speed.
Im watching the speed on the copy dialog transfer box when copying to a Network HDD and also when copying files with SCP between linux boxes its the same speed.
Just to clarify i have done this in two different locations, that means different gigabit routers and different cables and different destinations. The only thing in common is the laptop i have the USB adapter installed on.
Thanks for the additional details.
Many factors affect throughput when copying from a network location including:
- The size of files being copied (many small files tend to take longer than one large one).
- Congestion on the network from other connections.
- The speed of the data source drive, as well as whether any other disk activity is present here.
- Any slower network equipment between the USB3-E1000 and source system.
- What protocols are being used for the transfer.
Other factors may also be at play, but these are the most common ones.
To confirm if your USB3-E1000 is operating at full speed, it’s generally easiest to use a performance benchmark like iPerf to run some tests where the source isn’t data on a drive somewhere so you can avoid being limited by the factors above. When benchmarking the adapter, we connected one unit directly to another with a regular ethernet cable. A crossover cable isn’t required since the USB3-E1000 supports AutoMDIX or crossover cable detection. Since there’s no router in between, you’ll have to configure IP addresses manually for this test.
Softlayer has a great blog post on how to isolate what is slowing down network traffic: http://blog.softlayer.com/2011/using-…
Fundamentally, you’ll want to do a simple point-to-point test and then re-test as you add in each individual bit of network equipment in between the client and server. As you see speeds slow down after adding individual bits of equipment, you’ll know where your bottlenecks are if they are hardware. If you don’t see similar slowdowns just from adding network equipment, you’ll know the limiting factor isn’t hardware, but is something about either the data or the method of access.
Hope this helps, please let us know of any other questions!
Best wishes -