The ix4 leaves a bit to be desired when it comes to configuring its drives. You can choose only between JBOD, RAID 5 and RAID 10 volumes, which Iomega calls Without Data Protection, Mirror and Parity, respectively.
You can create only one volume with all drives in the volume and you have to delete all data from all shares before you change volume types. There also isn't the ability to do online expansion by swapping in one larger drive in at a time.
Figure 11: Volume creation options
The good news is that the drives are user serviceable. But Iomega doesn't support your pulling out all the drives and swapping in larger ones.
I didn't run our usual pull-a-drive RAID failure test because the one that Jim previously ran did so well.
The ix4 was tested with our standard test process. I used the latest 188.8.131.52472 firmware and ran tests with 1000 Mbps and 1000 Mbps with 4k jumbo LAN connections in JBOD, RAID 5 and 10 modes. Note that it took 24 hours to build an 8 TB RAID 10 array. I didn't time RAID 5 creation since the ix4 came configured in that mode. But I expect that a full RAID 5 build and sync would take the same time. As is typical, JBOD volume creation took only a matter of minutes.
I first checked backup performance to an Iomega UltraMax Pro Desktop Hard Drive configured in RAID 0 attached via USB 2.0. The test copies a 4.35 GB ripped DVD test folder that I use in the NAS Chart Vista SP1 file copy tests from the NAS to the attached drive. The results are summarized in Table 1 .
Backup Throughput (MBytes/s)
|USB - FAT32||14.9|
|USB - NTFS||4.4|
Table 1: Attached backup throughput test summary
The good news is that you can back up to an NTFS formatted drive. But the bad is that it won't be very speedy. FAT-formatted speed was about 5 MB/s slower than I've found with most current-generation NASes, with the highest performing products able to get over 20 MB/s.
I also tested NAS-to-NAS backup to the QNAP TS-239 Pro RAID 0 target that I have started to use as a standard rsync backup target and measured 14.9 MB/s. While I don't have a big comparison set, this was the same as I got with a Synology DS409+ backing up to a Synology DS409.
I ran the Vista SP1 filecopy test on a 10 GB iSCSI target configured above and measured 31.3 MB/s write and 34.8 MB/s read. This ranked the ix4 next to the bottom for write and at the bottom for read for the four, four-drive NASes that I have data for. (Say that three times fast!)
Write / Read
Figure 12 presents a summary of the benchmark tests with the write tests plotted. Write cache boost is quite moderate at the low end, with no values exceeding the speed of the Gigabit Ethernet connection. Once the 1 GB and higher file sizes are reached, write speed settles in at the mid 20 MB/s range for JBOD and low 20's for RAID 5 and 10.
Figure 12: Write performance benchmark summary
Read results for JBOD, RAID 5 and 10 are plotted in Figure 13 and show speeds around 40 MB/s for the 1 GB and larger file sizes.
Figure 13: Read performance benchmark summary
Average RAID 5 speeds over the 32 MB to 4 GB file sizes with cached results above 125 MB/s removed from the average measured 50.5 MB/s for write and 55.3 MB/s for read.