RAID Technology deserves Attention

RAID stands for Redundant Array of Inexpensive Disks, which is also read as Redundant Array of Independent Disks. It is simply an arrangement, where a number of less expensive hard disks are interconnected in order to amalgamate as a single disk, with high data storage capacity. This enables reliability and excellence in performance in the storage medium, deployed by the server markets or enterprises and that too at lower costs.

Note- Redundancy means, in a situation of failure in an array of disks, an availability of backup. In storage networking, the availability of a new disk on stand-by or a swappable disk, leads to Redundancy.

RAID levels

Level 0- In this level, the data gets stripped, which is splitting blocks of each file on multiple array of disks.

• (+)ve- Improves performance levels.

• (–)ve- one drive failure will lead to data loss, as there is no fault tolerance or redundancy offered in this level.

Level 1- disk mirroring is employed in this level and so there is also Duplexing feature in it.

• (+)ve- this level offers redundancy and fault tolerance.

• (–)ve- the need for another disk in deployment, makes it expensive. Storage capacity is equal to a smallest drive and downtime is observed, while changing an active drive, during failure.

Level 2- Striping at bit level is done, rather than block level and is rarely used.

Level 3- This level offers byte level stripping and that too with parity.

• (–)ve- multiple user requests are disqualified and so is not in practical use.

Level 4- This level implements level 0 block level stripping and also offers dedicated parity disk.

• (+)ve-In the case of any disk failure, the parity disk is used to build the replacement.

• (–)ve- The implementation of parity disks can create write bottlenecks.

Level 5- This is a good level of RAID storage in practical as it offers data striping at byte level and also offers correction in stripe error. Thus, this level offers performance excellence and also faults tolerance.

• (+)ve- performance increase is observed in storage array and full data redundancy is observed.

• (–)ve- Performance is narrowed during rebuilding of data and implementation is costly.

Level 6- In this level, the block level striping is implemented, along with the parity data distribution among the disks.

Level 7- This level is a Storage Computer Corporation trademark and offers caching to levels 3 and levels 4.

Level 0+1- This is a combination of a RAID level 0 and 1, where two RAID 0 levels are created with stripes and then the RAID 1 level is implemented for mirroring.

• (+)ve- Useful for replicating and sharing of data among the disks and data is redundant.

• (–)ve- effective data capacity is almost half and large number of drives are required.

Level 10- this is a vice versa of level 0+1 as the multiple level 1 disks are created and then the level 0 striping is implemented.

• (+)ve- performance is increased and data is redundant.

• (–)ve- effective data capacity is almost half and large number of drives are required.

Level 50- this is a combination of level 5+ level 0, where series of Raid 5 groups are striped in RAID 0 format.

RAID 53- Striping of data is done in Level 0 type and RAID 3’s virtual disk blocks are created.

• (–)ve- this offers higher performance and is a also high in cost.

RAID S- This is a proprietary level of EMC Corporation and offers striped parity, under Symmetrix Storage systems.

Note- (+)ve means advantages and (-ve) means disadvantages.

Storage requirements can be calculated through RAID Calculator.

Software RAID vs. Hardware RAID

There are two ways to deploy RAID technology, one is Software RAID in the sense of an operating system and the other is a hardware dedicated to it. In the case of a software , the CPU cycles of General computing are used for completing a RAID interface and this practically proves as a low cost RAID setup as just the implementation of a hard drive will do all the necessity. However, the performance of the whole system will be depleting and so the software RAID operation is now extinct, as the cost of hardware RAID setup has come down drastically.

In a Hardware RAID, a dedicated platform is created to reduce the load from the processor and this in-turn provides excellence in performance of array of storage. However, depending upon the RAID levels, the cost factor comes into play and this factor acts as selection criteria, from among the different RAID levels.

Conclusion

Depending on the version implemented, the benefits from RAID technology can be reaped. The cost factor is the driving force, while choosing RAID levels and depending on the needs of an organization, the technology can serve in accordance with the IT needs. This technology depends on the kind of hard drives, which are being aligned to form an array. Same brand and model of hard drives will do wonders, as they have the same capacity and performance levels. In the meantime, the capacity of hard drives, implemented according to the levels, will also shadow the performance of each drives in the array.

RAID Technology improvisations can be found here.