Thecus N4100PRO NAS Server ReviewNemo - July 19, 2010
» Discuss this article (2)
RAID Data Protection
Having multiple drives means more than just additional space for data. With four drives, the N4100PRO supports multiple RAID levels offering differing levels of redundancy. With two drives it supports RAID 1 mirroring and with three or more drives it offers RAID 5/6 redundancy. All four drives are required for RAID 10. The Thecus N4100 owners manual offers a good explanation of the different RAID levels. The N4100PRO also features RAID 5 + hot spare capability, which allows you to designate a drive as a spare drive that can automatically be added to the array in the event of a drive failure.
In order to test the ability of the N4100PRO to recover from a failed drive, I configured the unit with a RAID 5 volume using all four drives and also with a RAID 5 volume using three drives with the fourth drive designated as a hot spare. With its hot swap capability you can remove and replace drives without powering the unit down. During testing I simulated a drive 'failure' by simply removing the drive tray from the NAS. This results in an e-mail being generated alerting you of the failure, assuming you have enabled alerts in the system.
The first test began on the four-drive RAID 5 array where I removed drive 1 from the server. Checking the RAID Information page shows the RAID status as 'Degraded' and the Disks Used column only shows drives 2, 3 and 4. Since there is no hot spare, I manually replaced the drive in the server which caused the system to automatically begin the rebuild process. As you can see from the screen shot, this is a lengthy process with an estimated completion time of over seven hours. During the rebuild process, the data are still available but throughput is severely downgraded. Once the process is completed, the status returns to 'Healthy' and disks 1-4 again show up in the disks used column.
If, instead of adding all four drives to the array, you had used disks 1-3 for data and designated disk 4 as a hot spare, the Disks Used column would show '1,2,3,4' with the '4' in blue denoting it is a hot spare. In this test scenario, when I remove disk 1, the unit automatically begins the recovery process using disk 4. When the process completes, you'll see disks 2-4 in the Disks Used column and you can replace disk 1 and designate it as the new hot spare if needed.
The N4100PRO did an excellent job of protecting the data during a simulated drive failure, allowing the the unit to continue operating with a failed disk while rebuilding the array. Throughput rates in the NASPT test dropped to the 1.5-8.0MB/s range, but the server and its data remained available the entire time and did not require it to be powered off and no data was lost.
The N4100PRO offers you the ability to migrate from one RAID level to another. One of the benefits of this feature is you can expand a RAID 5 array, for example, from three disks to four. To do this you would edit an existing RAID array and click on the Migrate RAID tab on the RAID configuration screen. From there you click the RAID 5 -> RAID 5 radio button and click on Apply. Once the process completes you will see the extra space is now available.
Now that there is additional space available from the added drive, you can expand the size of the existing RAID volume using the Expand tab on the RAID Configuration screen. The screen shows the available space, in GB, as a percentage of the total space. You can then decide how much to allocate. Just be aware that the number shown is a percentage of the total array size, not of the available space. After you hit the Apply button you'll be greeted with several warning dialogs asking you to confirm the expansion process and then you're looking at a lengthy process as the additional space is added to the RAID array. Once complete, the additional space will show up in the Data Capacity column.
Thecus has made it very simple to add additional capacity to a volume. Coupled with the reliable auto rebuild we saw with the removal/replacement in the hard drive 'failure' tests, you could replace your drives one at a time with larger capacity units and then migrate and expand an existing array to take advantage of the extra space offered.