Have you ever owned a computer workstation or server equipped with a RAID array? They tend to be more common in Domain Controllers and Storage Servers, but are rarely found except in the most sophisticated of computer workstations. Of course people who use sophisticated 'precision workstations' with sophisticated software may demand such redundancy and performance from their hardware that they spec a RAID configuration.
I guess I am one of those people; I have almost always purchased Dell Precision Workstations as workhorses for my business, especially to run some of the sophisticated programs I use for data services. For years, my office also ran bank-after-bank of servers and storage arrays all using a variety of RAID technologies. Server RAID configurations tend to be hardware intensive and hard-drive heavy, but I always felt I got what I paid for when it came to redundancy and recovery capabilities.
What I didn't realize is that some of the more recent workstations, while expensive, don't feature the same level of redundancy and as such they are not as reliable as far as recovery. I recently suffered a RAID array failure and recovery was 'partly cloudy' at best. In fact even as I write this we are still working on recovering data, although we have reconfigured the array with replacement hard drives and controller. But enough of my sob-story, what I want to talk about is the value or lack of value of RAID.
So why do we buy a RAID computer? The number one reason is ‘redundancy’ and therefore ‘reliability’, but what happens when your RAID revolts? When the redundancy and reliability you expect fails? And why does it fail? These are questions we will look at in this mini-series on RAID or No RAID, that is the question.
Of course I realize that some people reading this may be asking, "what the heck is RAID, I thought that was the can of 'bug spray' underneath my kitchen sink?" So with that profound realization, I thought we should start off slow and in this first segment of this mini-series discuss the most common forms of RAID (and I don't mean 'ant & roach, flying insects, and wasp blaster).
RAID is an acronym that stands for redundant array of independent disks. RAID is a data technology that combines multiple computer disk drives for data redundancy and performance. With a RAID your computer's data is spread across multiple disk drives (an array) in what is referred to as RAID levels, depending on the desired level of redundancy and performance. Each RAID level provides a different configuration of redundancy, performance, reliability, capacity and availability.
So in this first article let's look at the 3 most common, which also progress in oldest to most recent (of the three).
RAID 0
-
RAID 0 is a configuration in which the system data is split-up and written across all the disk drives in the array. A minimum of 2 hard disks are required, but a RAID 0 configuration can accommodate as many drives as are available to the array.
Because this configuration maybe combined with multiple RAID controllers, ideally one per hard disk, this configuration offers maximum performance but no redundancy, since there is no overhead caused by parity controls. RAID 0 is NOT fault tolerant, so if one drive fails in a RAID 0 configuration, then data will be lost.
RAID 1
-
RAID 1 is a configuration in which all system data is written to both a primary and mirrored drive (or set of primary and mirrored drives). This configuration requires a minimum of two drives. Should one drive fail, the RAID Controller uses either the primary or mirrored drive for data recovery and operations should continue until the failed drive can be replaced.
RAID 1 is not noticeably slower in performance than a single drive. The downfall is that the storage capacity is equivalent to only the capacity of the primary drives, even though the purchase price has been doubled. RAID 1 is considered ideal for mission critical applications and small servers.
RAID 5
-
RAID 5 is probably the most common RAID configuration in use, it requires a minimum of 3 disk drives but can use significantly more. In this configuration data is written across all the drives with 1 drive holding a parity check sum of the data on all the drives. The parity data is spread across all the drives in order to provide reliability by allowing the controller to recalculate the data when one drive is lost.
Raid 5 is an extremely complex technology that combines efficient storage with above average data security; however, performance is generally slower than with other RAID configurations. This form of RAID is commonly used in performance workstations and servers that have limited hard drive capacity in order to maximize storage while simultaneously providing a degree of redundancy.
This is pretty boring stuff to all but the most tech-savy (who already know it), so I think that is probably where I should stop boring you for today. Next time we will look at RAID 6 and RAID 10 configurations, along with both software and hardware RAID technologies.