Week 3: Clear - Importance of Redundancy

In business and enterprise computing, availability of data and information is probably the most important criteria of information technology.  To ensure the availability of data and services, the hardware and configurations must be redundant in all aspects.  Redundancy can be accomplished in multiple ways.  There are redundant hard drives, power supplies, network interface cards (NIC), network switches and routers, just to name a few.  Unfortunately, there are components that cannot be redundant, which are motherboards, hard drive controllers, operating systems, and software.  To aid in non-redundant aspects, servers can be setup to do clustering, virtual machines, and network load balancing.  Hardware failures can occur without any warning and must be planned for based on the criticality of the host and data.

Hard drive redundancy is accomplished using a technology called Redundant Array of Independent Disks, also known as RAID.  RAID is a storage technology which utilizes multiple disks to create a virtual storage volume.  This storage volume provides fault tolerance and greater throughput or speed.  This can be accomplished via a hard drive RAID controller or using software RAID.  The two most used RAID configurations are RAID level 1, which is essentially mirroring, and RAID level 5, which utilizes block-level striping with dedicated parity.

The following URL provides several visuals for each RAID configuration. Reference URL: http://www.ecs.umass.edu/ece/koren/architecture/Raid/basicRAID.html

Picture taken from: http://farm6.staticflickr.com/5258/5528362934_4be047e5b5.jpg