| RAID Type | Description | Disks needed |
| RAID 0 | Two or more hard drives that are bound together into a single virtual disk volume. Data is written to the stripe set across the array instead of one disk at a time. This provides RAID 0 with a definite performance improvement over non-RAID storage simply because you’re drastically improving data transfer per read or write cycle. RAID 0 is not a fault tolerant system. If one participating disk fails, you lose all data in the striped volume across the entire disk set. Always back up RAID 0 volumes religiously! | |
| RAID 1 | RAID 1 sets are called mirrored volumes because you mirror every disk I/O operation on the primary disk member to a second, separate disk called the mirror disk. Therefore, you implement RAID 1 when data recovery is foremost on your mind. For example, many system administrators will mirror the system drive of their servers to recover from a boot problem with minimal downtime. Depending on your RAID controller, you can configure the mirrored drive to automatically assume the role of primary member if that primary disk goes offline. For software RAID like what windows has, manual administrator intervention is always required. | |
| RAID 5 | It is called a stripe set with parity. This RAID type attempts to blend the benefits of RAID 0 with the data redundancy of RAID 1. You need at least three separate hard disks to build a RAID 5 set due to the redundant parity data that is required. You also lose one disk’s worth of data. The good news, though is that a single drive failure won’t destroy the data on the RAID 5 volume. You can simply replace the failed disk and let the RAID controller rebuild the RAID 5 volume by using the stored parity data. It cannot recover from a multi-drive failure. | At least three disks needed |
| RAID 10 | It can survive multiple drive losses as long as no mirror partner loses all its members. |