Raid Stands for Redundant Array of Independent Term Paper

Pages: 4 (1741 words)  ·  Bibliography Sources: ≈ 6  ·  File: .docx  ·  Level: College Senior  ·  Topic: Education - Computers

RAID stands for Redundant Array of Independent or Inexpensive Disks, a type of Computer Disk drives or storage media that uses two or more drives in association for fault tolerance and performance. The use of RAID disk drives are normally made on servers in a computer network environment. (RAID: (www.webopedia.com) Fault tolerance is the capability to react smoothly and thereafter recover from a sudden hardware or software breakdown. Several stages of fault tolerance exist, the lowest stage being the capability to continue running the system in the event of power disruption. A lot of fault-tolerant computer systems emulate every function which indicates every function is executed on two or more identical systems, such that in the event of breakdown of a particular system the other one can continue operation from that stage onwards. (Fault Tolerance: (www.webopedia.com)

The fundamental concept of RAID was to unite multiple small, cheap disk drives into an array of disk drives that performs far in excess of a Single Large Expensive Drive (SLED). Moreover, this collection of drives seems to the computer system as a unitary logical storage unit or a disk drive. In the event of failure, the concept of Mean Time between Failure -MTBF of the array of disk drives will be equivalent of the individual disk drive, divided by the number of drives present in the array. Due to this situation the MTBF of an array of disk drives will be very low for a lot of application needs. Nevertheless, disk array can be rendered fault-tolerant through redundantly storing information in a number of methods. (What is RAID)Buy full Download Microsoft Word File paper
for $19.77

Term Paper on Raid Stands for Redundant Array of Independent Assignment

Central to RAID is "stripping," a procedure of joining multiple drives into a single logical storage unit. Striping is the procedure of partitioning the storage space of each drive into stripes that might range from at the minimum of one sector which is 512 bytes up to several megabytes. Thereafter these stripes are interleaved in a round-robin manner; such that the total space is constituted alternatively of stripes from each drive. As a result, the storage space of the drive is mixed up like a pack of cards. The nature of application environment, Input-Output (I/O) or data intensive, governs if large or small stripes must be employed. Nearly all of the multi-user Operating System -OS like NT, UNIX and Netware are compatible with overlapped disk I/O' operations throughout multiple drives. but, to maximize throughput for the disk subsystem, the I/O load should be balanced though all the drives such that could remain busy to the extent possible.

In case of a multiple drive system without stripping, the disk I/O load is not at all evenly balanced. It might so happen that certain drives must be having data files that are repeatedly accessed and some drives will be sparingly accessed. In case of I/O intensive settings, peak performance is achieved through stripping the drives in the array with stripes sufficiently big such that every record comes within a particular stripe. This makes sure that the data and I/O will be uniformly spread across the array permitting every drive to perform on a different I/O operation, and therefore maximize the number of concurrent I/O operations that can be executed by the array. In system settings where there are more data and single user systems that access bigger records, small stripes, the length is normally one 512-byte and can be utilized such that each record will cover throughout every drives within the array, with every drive storing a portion of the data from the record. This results lengthier record accesses to be executed rapidly, as the data transfer happens in parallel on multiple drives. (What is RAID)

Regrettably, small stripes negate multiple overlapped I/O operations, as every I/O will usually entail every drive. But OS like DOS which does not permit overlapped disk I/O will not be affected negatively. Applications like streaming video / audio, medical imaging and data capture, which utilize lengthier record accesses, will achieve optimum performance with small stripe arrays. A possible limitation to employing small stripes is that synchronized spindle drives are needed so as to keep performance from being degenerated when smaller records are accessed. In the absence of synchronization, every drive inside the array will be at varied arbitrary rotational positions. As an I/O cannot be finished till each drive has accessed its part of the record, the drive which takes the longest will find out the time when I/O complete. Greater the number of drives in the array, the greater the average access time for the array approaches the worst instance of single-drive access time. Synchronized spindles make sure that every drive in the array reaches its data at the identical time. The access time of the array will therefore be identical to the average access time of the individual drive quite than approaching the worst case access time. (What is RAID)

Types of RAID:

i) I/O controller (IOC) based RAID:

This type of RAID utilizes the processors and memory resources of the I/O controller to run the RAID engine. Even though it appears identical to the IOP-based RAID, IOC-based RAID has restricted characteristics and workability as it shares the restricted processor and memory resources of disk I/O controllers. Due to its minimum footprint and being cheap, IOC-based RAID is generally implanted on the server platform.

A ii) Driver-based RAID:

In this case, software implementation is unified in the driver of a particular disk controller. The driver has the programming code to execute the RAID engine within the OS setting and the BIOS execute the RAID engine in the prior to OS (startup) settings. In the OS, this method appears identical to IOP and IOC-based RAID to the effect that real physical drives comprising the RAID volume are invisible to the OS. Nevertheless, driver-based RAID is dependent wholly on the resources of the system processor and memory for RAID running and impacts the system performance in case of high CPU utilization settings.

A iii) OS-Based RAID:

In this category of software, RAID is generally executed as a filter driver bundled within the OS, utilizes the processor and memory of the host system to execute the RAID engine. OS-based RAID is not dependent on disk controller type and utilizes disk controller for disk I/O. (Comparing RAID Implementation Methods)

Levels of RAID:

RAID consists of nine levels and also a non-redundant array which is RAID-0.

RAID -0: This method has stripping however no redundancy of data is there. It provides the optimum performance however there is no fault-tolerance.

RAID -1: This category is also called as disk mirroring and comprises at the minimum two drives that keep identical records of the storage data. No instances of striping occur. Read performance is enhanced as any one of the disks can be read concurrently. Write performance is identical to the individual disk storage. RAID-1 generates the most excellent performance and the most excellent fault-tolerance in case of a multi-user setting.

RAID-2: This category of employs stripping throughout disks with certain disks keeping in memory, error examination and remedying ECC information. It does not have any edge over RAID-3.

RAID-3: In this case stripping is employed and dedicates a single drive to storing parity information. The embedded error checking -ECC information is employed to find out errors. Recovery of data is done by taking into account the exclusive or -XOR of the information stored on other drives. As an I/O operation tackles all the drives concurrently, RAID-3 cannot overlap I/O. Due to this RAID-3 is ideally suited for individual-user systems having lengthier record applications.

RAID-4: This category utilizes large stripes which indicate one can read records from any individual drive. This permits one to take benefit of overlapped I/O in case of read operations. Because every write operations have to update the parity drive, no overlapping happens. RAID-4 does not offer any benefit compared to RAID-5.

RAID-5: In this type of RAID it comprises a rotating parity array, therefore taking into account the write restriction in RAID-4. Hence, every read and write functions can be overlapped. RAID-5 keeps in memory parity information but not the redundant data however; parity information can be utilized to remake the data. RAID-5 needs at the minimum three and generally five disks for the array. It serves as the most excellent in case of multi-user systems wherein performance is not vital or which do less write operations.

RAID-6: This type of RAID is identical to RAID-5 however comprises a second parity scheme which is distributed throughout several drives and hence gives tremendously high error and malfunction of drive tolerance.

RAID-7: This category contains a real-time implanted OS as a controller, caching through a high-speed bus and other features of a non-networked computer.

RAID-10: This category provides an array of stripes wherein every stripe is a RAID-1 array of drives. This offers more performance compared to RAID-1 nevertheless at much increased cost.

RAID-53: Under this category an array of stripes is provided wherein every stripe is a RAID-3 array of disks. This… [END OF PREVIEW] . . . READ MORE

Two Ordering Options:

?
Which Option Should I Choose?
1.  Buy full paper (4 pages)Download Microsoft Word File

Download the perfectly formatted MS Word file!

- or -

2.  Write a NEW paper for me!✍🏻

We'll follow your exact instructions!
Chat with the writer 24/7.

Redundant Array of Independent Disks Essay


Stand-Your-Ground Laws to Persuade Essay


Independent Auditor Has Been Playing an Important Term Paper


Taking a Stand Research Paper


Independent Study Program Term Paper


View 200+ other related papers  >>

How to Cite "Raid Stands for Redundant Array of Independent" Term Paper in a Bibliography:

APA Style

Raid Stands for Redundant Array of Independent.  (2004, November 27).  Retrieved February 27, 2020, from https://www.essaytown.com/subjects/paper/raid-stands-redundant-array-independent/4707

MLA Format

"Raid Stands for Redundant Array of Independent."  27 November 2004.  Web.  27 February 2020. <https://www.essaytown.com/subjects/paper/raid-stands-redundant-array-independent/4707>.

Chicago Style

"Raid Stands for Redundant Array of Independent."  Essaytown.com.  November 27, 2004.  Accessed February 27, 2020.
https://www.essaytown.com/subjects/paper/raid-stands-redundant-array-independent/4707.