After trying out ZFS on a handful of fitful USB keys I pulled out the AMS Venus T5 a 5 drive SATA RAID enclosure with an eSATA port for plugging into the computer. The Venus T5 was outfitted with 5 x 7200RPM 1TB Seagate Drives for the test. I was curious as to what the difference would be
RAID 0 setting has them all striped together. The Single Drive results are an average from all 5 drives. To keep in with the previous post I still used date +%s for the timings instead of using the more accurate time command.
Standard Test (5 cycles):
| Single Drive | Hardware RAID 0 | Software RAID 0 | ZFS | |
|---|---|---|---|---|
| Sequential Read Speed: | 71.542 | 68.170 | 66.358 | 678.275 |
| Sequential Write Speed: | 63.107 | 68.043 | 64.925 | 338.702 |
| Random Read Speed: | 17.203 | 21.308 | 28.053 | 276.802 |
| Random Write Speed: | 29.525 | 29.634 | 46.104 | 505.464 |
Extended Test (5 cycles):
| Single Drive | Hardware RAID 0 | Software RAID 0 | ZFS | |
|---|---|---|---|---|
| Read Speed: | 100.501 | 113.045 | 125.893 | 248.304 |
| Write Speed: | 90.517 | 97.360 | 101.469 | 299.403 |
As seen in the ZFS test on the USB keys it’s quite obvious that QuickBench 4.0 is useless in benchmarking a ZFS formatted drive in comparison to other formatted drives. ZFS isn’t supported so it isn’t surprising.
As a continuation of a more consistent test we then copied a 903MB file (Xcode 2.5) and a folder of of 1500 0 byte files numerically named.
date +%s; cp ~/Desktop/SpeedTest/xcode.dmg DESTINATION; date +%s;
cp ~/Desktop/SpeedTest/*.txt DESTINATION; date +%s
| Single Drive | Hardware RAID 0 | Software RAID 0 | ZFS | |
|---|---|---|---|---|
| Large File Time: | 0:36 (~25MB/sec) | 0:31 (~29MB/sec) | 0:29 (~31MB/sec) | 0:41 (~22MB/sec) |
| Small File Time: | 0:01 | <0:01 | <0:01 | <0:01 |
With much better hardware the actual throughput information is a lot more interesting.