Do you have a super fast network, but it still feels like it takes forever to copy or retrieve a file from an AirPort Disk? It may be possible that your network is not optimized to provide the highest throughput that it can offer ... but how do you get a objective measurement to compare against?
There are a few things to take into consideration, namely:
- The throughput between your Mac and the AirPort Extreme Base Station (AEBS) or Time Capsule (TC), and
- The throughput between your Mac and the USB HDD attached to the AEBS or TC.
For comparison I did a couple of simple tests for my network. In my case, I tested throughput using an Mid 2011 Intel Mac mini and a 1st generation 802.11ac TC. Attached to the TC is a Western Digital Passport USB HDD. The basic network path between them is: Mac mini > Ethernet Switch > Ethernet Switch > Cisco Router > Ethernet Switch > TC > USB > WD Passport. (Note: All network connections are via Gigabit Ethernet, except the connection between the TC and the HDD which is via v2.0 USB. Substituting wireless connections, at any junction, will result in an overall reduction in throughput.) As you can imagine, you can have literally hundreds of other network configurations and each will be unique for their overall throughput characteristics.
Measuring network throughput between the Mac mini and the TC
Method 1 - Using IPNetTunerX
I first measured the data link rate from the Mac mini to the TC by using IPNetTunerX. I used iPNetTunerX's Link Rate test tool. I did five runs and the average throughput was found to be 925.0 Mbps (115.6 MBps). Not bad as this is near the maximum possible bandwidth of a Gigabit Ethernet connection.
Measuring throughput between the Mac mini and the AirPort Disk
Method 1 - Using Net Monitor
To measure throughput to the WD Passport, I first created a 1 GB disk image file using the Disk Utility. I then copied this file to a folder on WD HDD. To get the throughput reading I used Net Monitor. I could have easily used OS X's Activity Monitor for this as well. I did five runs of copying this file and the average throughput was found to be Read: 445.6 / Write: 256.0 Mbps (Read: 55.7 / Write: 32.0 MBps).
Method 2 - Using AJA System Test
A second way to run a throughput test to the AirPort Disk is to use AJA System Test. System Test will automatically create a temporary file and perform read/write tests, then delete the file when done. Again, I did five runs transferring a 1 GB file. Results: Read: 464.0 / Write: 296.0 Mbps (Read: 58.0 / Write: 37.0 MBps)
Method 3 - Using Disk Speed Test
For the final test runs I used Blackmagicdesign's Disk Speed Test. I did five runs transferring a 1 GB file. Results: Read: 468.0 / Write: 304.0 Mbps (Read: 58.5 / Write: 38.0 MBps)
Conclusion
As I expected the throughput to the AirPort Disk would overall be significantly lower than that over the network connection alone, as they would be additive. All in all, it took less than a minute to transfer a 1 GB file.
For comparison, let's look at some other baseline test results:
- Test configuration #1: Mid 2011 Mac mini, internal Macintosh HD
- Results: Read: 616 / Write: 744 Mbps (Read: 87 / Write: 93 MBps)
- Test configuration #2: Mid 2011 Mac mini, 2GB WD HDD in NewerTech MiniStack enclosure directly connected to Mac mini by FireWire 800
- Results: Read: 612 / Write: 616 Mbps (Read: 76.5 / 77 MBps)
- Test configuration #3: Mid 2011 Mac mini, 2GB WD HDD in NewerTech MiniStack enclosure directly connected to Mac mini by USB 2.0
- Results: Read: 275.2 / Write: 236 Mbps (Read: 34.4 / 29.5 MBps)
You can perform similar tests to determine if the issue is the connection to the AEBS/TC or to the external USB HDD that you are using. These two values should not be off more than 50-55%. If they are, I would suspect the USB HDD. If both are extremely low, then it would indicate the connection to the AEBS/TC will require further troubleshooting.
Replies