Today consumers are accustomed to having direct and uninterrupted access to online services. As a service manager, you have the ability to choose between many kinds of systems and infrastructures for critical online applications. One of the features that the administrator needs to consider is RAID, but more important is the choice of the RAID setup structure.
RAID, originally from the redundant array of independent (or inexpensive) disks, is a disk subsystem that keeps the files on multiple disks and succeeds higher speeds, fault tolerance or even 2 together.
There are two ways to implement a RAID layout: Hardware or Software.
Managing a hardware layout is done by a dedicated hardware controller on which the disks are connected. The controller does its own calculations as it has a microprocessor built into it to avoid overloading the system. Nowadays, however, today's processors have evolved so much that this privilege is virtually unnecessary. However, a controller provides an extra level of security with the Battery Backup Unit (BBU), as in the event of an unexpected power failure, no records are lost.
The RAID software is through the operating system and is the most cost-effective way to implement it as it does not require additional hardware resources.
Below we can see one of the simplest RAID devices used, while in our next article, we will analyze even more provisions.
RAID 0 (Disk striping):
The RAID 0 device shares the data on 2 or above drives, known as striping, allowing for greater data throughput. Each file is read by multiple disks by combining the speed of all media together. However, it does not provide some kind of error tolerance as it does not keep copies of the files nor parity info (more about parity in next article). Both or more disks appear as a partition, so if one is spoiled, it spoils the entire array, leading to data loss. This layout is usually used to record live streams where speed comes first while reliability is second.
The sequel to the article ... in the link below.