Usually some computer motherboards already support with raid10. Raid10 using a set of disks with striping at the block level without redundancy. So just do the striping blocks of data into multiple disks. Excess of this level has already include access to several blocks and can be done in parallel so it can be faster while sending data. However lack of access per block, just like there is no improvement, less reliability in the absence of data backups with redundancy.
Sunday, December 11, 2011
Various methods to improve the performance of the disk
There are various methods used to improve the performance and capability of the disk. The methods are usually used to improve performance, it will involved many disks as one storage unit. Each block of data is separate into some sub block, and is divided into disks (striping). When there are data transfers, the disks will work in parallel, and increase the transfer speed in reading or writing data. Combined with the rotation synchronization on each disc, then the performance of the disk can be increased. This method is known as RAID - Redundant Array of Independent Disks. Beside that RAID performance problems may also increase the reliability of the disk by using additional disks to store the parity block or as a mirror of the data on the disk.
Usually some computer motherboards already support with raid10. Raid10 using a set of disks with striping at the block level without redundancy. So just do the striping blocks of data into multiple disks. Excess of this level has already include access to several blocks and can be done in parallel so it can be faster while sending data. However lack of access per block, just like there is no improvement, less reliability in the absence of data backups with redundancy.
Usually some computer motherboards already support with raid10. Raid10 using a set of disks with striping at the block level without redundancy. So just do the striping blocks of data into multiple disks. Excess of this level has already include access to several blocks and can be done in parallel so it can be faster while sending data. However lack of access per block, just like there is no improvement, less reliability in the absence of data backups with redundancy.
Subscribe to:
Posts (Atom)