There are many benefits of using a Redundant Array of Independent Disks (RAID) ranging from, from an increase in data read/write speed through to data redundancy. All RAID levels are a compromise between hardware requirements, read/write data throughput and the security of your files.
Budget will usually be the main driving factor in determining which RAID level is most appropriate to suit your requirement. When there are no budgetary constraints data security should be one of the important factors in the decision. It is important in all cases not to believe that just because you have a RAID array you don’t need a backup. A failure to provide an adequate backup solution will usually result in requiring RAID data recovery services at a later date.
RAID 10 Gives 100% Redundancy
A RAID 10 array stripes data across a set of disks arranged in mirrored pairs, which requires twice the number of drives, for the desired capacity. Each mirrored pair provides full redundancy, but the failure of any single drive could be closely followed another, common in RAID arrays where the disks operate in the same environment. If both disks in a mirrored pair fail at the same time, the RAID will fail, meaning that despite the full redundancy, there is still an element of risk.
Data throughput in a RAID 10 can often be considerably higher than other RAID systems as there is no need to calculate parity for each write operation. The hardware in a RAID 10 array is often set up to take the data read from the fastest responding drive in a mirrored pair. In theory it is possible for a RAID 10 system to continue operating with 50% of the drives failed, providing only one disk in each mirrored pairs fails, allow such action would be to run an unnecessary risk to the security of your data. For high availability servers, such as those running Exchange and SQL databases RAID 10 is a common choice.
RAID 5 Capacity Over Speed and Redundancy
In a RAID 5 array the data is striped across each drive, with a single drive in each data slice containing the parity information, which in the event of a drive failure can be used to reconstruct the missing data. Effectively only a single drive is used for redundancy rather than a doubling of the disks, thereby allowing for much larger data volumes for a given set of hard disks.
It is possible for a RAID 5 array to run in degraded mode following the failure of a single drive. This does however lead to reduction in performance while putting your data at imminent risk should another disk fail. Despite this downside compared to a RAID 10 system, RAID 5 is still one of the most commonly used RAID array architectures.
RAID Data Recovery
Although RAID 10 arrays consist of mirrored drive pairs, they are still sometimes seen for data recovery when both drives in a single drive pair fail. It offers double the chance of recovery, thereby increases the success rate for data recovery from RAID 10 arrays. If the failure of a drive in a RAID 10 mirrored pair is ignored it may hold out-of-date data which may reduce the chances of a completely successful data recovery.
RAID 5 arrays may have an inherent higher level of risk but the success rates for data recovery are also very high. This is because drive failures are rarely severe enough to cause the loss of large areas of the data volume.
The use of redundancy will always be better than none again, so the choice really comes down to budget, and how much risk you’re willing to take with the overall integrity of your data. This needs to be weighed against the possible financial harm your company would face, even for a temporary loss of data access.