Emergency Server Recovery Services

24 Hour Server Recovery Helpline Tel: +44 (0)1865 469468

RAID 5 or RAID 10

The use of a RAID (Redundant Array of Independent Disks) offers a range of benefits, including increased data transfer speeds, redundancy and the ability to create very large data volumes. The benefits for each RAID configuration are a compromise between data security, data transfer speeds and overall hardware required.

If you have a limited budget it will put a constraint on the overall size of the RAID array and may compromise the priority requirement. If there are no budgetary restrictions data security through redundancy should be a very high on the list of requirements. Another very important consideration is to understand that RAID redundancy is not a replacement for a backup strategy otherwise RAID server data recovery will almost certainly be required in the future.

RAID 10 Full Redundancy

In a RAID 10 array data is striped across a set of disks arranged in mirrored pairs, which necessitates double the number of drives for the required capacity. While this provides full redundancy with the ability to rebuild any drive following a disk failure, the possibility of another drive failing within 24 hours is very high. If the mirrored hard disk drive in a pair should either fail, or contain unreadable bad sectors, this will cause the RAID to go offline. Even with 100% redundancy, which provides the highest level of data security, there is still a risk.

Without the need to recalculate parity, RAID 10 can in many instances provide faster read and write times. The hardware in a RAID 10 server is often set up to take the data from the disk which responds the fastest. It is in theory possible for a RAID 10 to run with 50% of the drives in a failed state, providing a mirrored pair does not fail. Continuing to operate with even one drive failed without rebuilding puts the integrity your data at serious risk. RAID 10 configurations most commonly used for high availability servers, such as those running Exchange and SQL databases.

RAID 5 Provides Higher Capacity

In a RAID 5 server the data is striped across the drives, with one drive in each data slice containing the parity information. This parity information can be used to reconstruct the data for a single failed drive which is being rebuilt. Therefore only the capacity of a single drive is used for redundancy, allowing for much larger data volumes using the same number of hard disk drives.

It is possible for a RAID 5 array to run in degraded mode if only a single drive fails. This however causes a performance hit, along with putting your data at imminent risk due to the failure of another drive. If a second drive fails this will cause the RAID to fail and go offline. RAID 5 is however still one of the most commonly used RAID array architectures, as it provides a good compromise between capacity, data security and transfer.

RAID Data Recovery

RAID 10 arrays are still sometimes seen for data recovery despite the use of mirrored pairs. Providing failures have not been ignored, following the failure of one drive in a mirrored pair which could hold out-of-date data, RAID 10 offers a double chance at recovery the data from each slice of the RAID, giving extremely high rate of success for data recovery.

While RAID 5 arrays have a higher level of risk attached, the success rate for data recovery is also very high. This is because it’s rare for the drive failures to be so severe that large sections of data cannot be recovered.

The use of redundancy is always a better option than none at all, so the choice really comes down to your budget and the level risk you’re willing to take with the overall integrity of your data. It is always important to consider the possible financial harm your company would face if you suffer even a temporary loss of access to your data.

« «
Copyright © 2015 DiskEng Emergency Server Recovery - All Rights Reserved