As I do with all NAS testing, I used IOzone to check out the file system performance on both ReadyNAS models (the full testing setup and methodology are described on this page). Since Infrant obviously recommends using the ReadyNAS with RAID 5 protection, that's how I tested the 600. The X6 was shipped with only two drives connected so that I could see how the auto-expansion capability worked, so that's what its test results reflect.

I tested with each model and computer running iozone connected via a 10/100 switch, then moved them to a NETGEAR GS108 10/100/1000Mbps switch for gigabit-speed testing. I used the GS108 engineering sample that I recently reviewed that supports jumbo frames up to 9k, so I was also able to exercise both boxes in that mode. The iozone computer connected via an Intel PRO/1000 MT Desktop gigabit Ethernet adapter, which was set to auto-negotiate its connection speed.

NOTES:
• Firmware version tested was V2 p1c2 1_00a121for both 600 and X6
• Default settings included disk write caching enabled, journaling enabled, but full journaling disabled
• Keep in mind that the maximum raw data rate for 100Mbps Ethernet is 12500 kBytes/sec and 125000 kBytes/sec for gigabit

Figure 18 shows a 3D surface plot of the ReadyNAS 600 write performance using file sizes from 64 kBytes to 128MBytes and record sizes from 4 kBytes to 16MBytes. Note that the plot is oriented so that larger file sizes are closer to the front.

Figure 18: ReadyNAS 600 Write performance - 100Mbps
(click image to enlarge)

The transfer rates well in excess of 100M Ethernet speeds show the effects of write caching. I'm not showing the X6 results because they were essentially the same. But I am showing the 100M RAID 5 results for the Buffalo TeraStation in Figure 19, because they are significantly lower.

Figure 19: Buffalo TeraStation Write performance - 100Mbps RAID 5
(click image to enlarge)

Given that the caching effects dominate the plot, the difference in gigabit-connected performance isn't obvious. But if you compare the lower regions at the front of Figures 18 and 20, you can see that the "plateau" in Figure 20 is somewhat higher.

Figure 20: ReadyNAS X6 Write performance - 1 Gbps
(click image to enlarge)

I switched to showing the X6 plot this time, because it shows the effects of kicking in 4k jumbo frames a little better when compared to Figure 21.

Figure 21: ReadyNAS X6 Write performance - 1 Gbps, 4k jumbo frames
(click image to enlarge)

Once again, because of caching, jumbo frames don't make a huge effect in peak write throughput. But they do help out as caching effects fade away with larger file and record sizes. I didn't test the TeraStation for jumbo frames, so can't present any data for it.