This video will define what RAID is, discuss the characteristics, advantages and disadvantages of various RAID levels. We will also review some key RAID terms.
This video will define what RAID is, discuss the characteristics, advantages and disadvantages of various RAID levels. We will also review some key RAID terms.
RAID is an acronym for redundant set of independent or inexpensive disks. RAID can be implemented through hardware or software. The technology is used to write reliability and redundancy to data and to enhance performance.
For software RAID, RAID arrays are created by combining partitions and disk spaces in the operating system environment. For software RAID hardware RAID controller is not necessary.
In order to set up a hardware RAID a hardware controller and several drives are required. In a hardware RAID the physical drives are transparent to the operating system and the operating system will only see the RAID virtual disks.
For a hardware RAID, RAID utility is usually provided to configure the array, normally called the RAID bios, can be accessed by key sequence during post. Hardware RAIDs are recommended for critical applications and provide easier rebuild of data stripes in case of failure.
The supported RAID levels on Dell servers are 0, 1, 5, 10, and 50. Depending on the controller capabilities. Each level combines drives at different arrangements in order to improve reliability and performance. RAID 0, 10 and 50 arrays provide faster performance, because multiple disks can be accessed simultaneously. RAID 1, 10, 5 and 50 arrays provide data protection with disk mirrors or parity.
Characteristics of RAID 0 include that it is not a true RAID since it offers no redundancy. With the RAID 0, data is distributed evenly across one or more drives and requires one or more drives to implement. Recommended applications for RAID 0 include audio and video streaming and editing, Web servers, graphics design, and applications that require high speed data access and storage.
But does not require redundancy. RAID 0 advantages include high speed data access, improved input and output performance, easy implementation and a low overhead because no parity calculation is involved. One disadvantage to RAID 0 is that it has no redundancy. RAID 1 characteristics include mirroring which writes an exact copy to two different drives. It requires two drives to implement.
Applications suggested with use with RAID 1 include financial applications, small databases, applications that require redundancy and faster read rates, and entry level systems that require redundancy but only two drives. Advantages to RAID 1 include 100 percent redundancy, if one drive fails an exact copy remains.
Disadvantages to RAID 1 include that only 50 percent of the capacity of the drives are being used. Characteristics of RAID 5 include striping with distributed parity. Parity allows RAID 5 to recover from single drive failure. It requires a minimum of three drives to implement.
Recommended applications for use with RAID 5 include file and application servers. Database servers, Web e-mail and news servers. Internet servers. RAID 5 is considered the most versatile RAID level. Advantages with RAID 5 include redundancy through a parity stripe, using a minimum drive overhead for parity, and possible simultaneous read, multiple drives.
Disadvantages with RAID 5 include performance being impacted when one drive is lost, and write speeds affected due to parity overhead. Raid 10 is also referred to as a striped set of RAID 1s. RAID 10 characteristics include being implemented as a stripe array with RAID 1 segments.
RAID 10 requires four or more drives to implement. Recommended applications for use with RAID 10 include database server requiring high performance and fault tolerance. Small databases and sites that would have gone with RAID 1 but need additional performance.
Advantages to RAID 10 include high input/output rates by striping RAID 0 segments. Under certain circumstances, RAID 10 can also sustain multiple simultaneous drive failures. Disadvantages for RAID 10 include it being very expensive with a high overhead and lower sustained performance as drives must move in parallel. Also, limited scalability at a high inherent cost.
RAID 50 is also referred to as a striped set of RAID 5 arrays. Characteristics include that it requires six or more drives to implement. RAID 50 allows recovery from single drive failure and RAID 5 sets but not to drive failures within a set. Recommended applications for use of RAID 50 include that it may be good for sites looking up RAID 5 but need additional performance boost.
RAID 50 advantages include that it is the highest level of redundancy and performance. High data transfer rates are achieved with RAID 5 array segments. High input/output rates for small requests are achieved with RAID 0 striping. Disadvantages to RAID 50 include that it is very expensive to implement.
All disk spindles must be synchronized limiting the choice of drives. Failure of two drives in one of the RAID 5 segments renders the whole array unusable. In this section we'll outline why RAID should not be considered a replacement for a back-up solution.