Raid0 : same as Raid1, perfectly split node. Files are written as nodes sequentially.
Raid1: one or more ssd.s' partitioned into 2 parts including partitioning spare area into 2 parts.
That represents a node (a dot) in space which means files are written as points in space independently. Second partition is set for references. A node size of a block which is 512k.
Raid2: means having one or more ssd's partitioned into 3 parts as well as partitioning Spare area into 3 parts. That is to have a line along x axis. More, to have files to line up on a single axis.
Raid3,4,5:: is to have files to be placed on layers on a single plane in 3-D.
.
.
Raid6: an ssd or more partitioned into 7 parts as well as the spare area. So we can store files in 3-D space. More, that is to create 3-D storage area.
The rest to come soon. Need to drink a glass of white coloured liquid.
Wear levelling on Raid6 using 2 ssds subdivided identically: Create 7 partitions on each ssd as well as 7 partitions in spare area and stack them together. That will generate 2 cubes of storage area in 3-D space. One to utilise as a storage area and the other to be used as additional spare area. As soon as the valid area and spare area worn out fully, then each file will be sent to next ssd's valid area as their space wear out. Once all the blocks are dead on the first ssd, the files would have been transfered to the next ssd as they were on the initial see. Since the first see is totally worked out, the second can be the main storage area. And we can replace the first see with a new see on a clean state raided6 (that is partitioning valid and spare area with 7 partitions each). And thus we can always have the storage ssd'in fin a state of free of dead blocks. This can be repeated as the storage ssds keep dying on you and have a clean see which is no dead blocks.
Timeout for a glass of white coloured liquid.
Raid1: one or more ssd.s' partitioned into 2 parts including partitioning spare area into 2 parts.
That represents a node (a dot) in space which means files are written as points in space independently. Second partition is set for references. A node size of a block which is 512k.
Raid2: means having one or more ssd's partitioned into 3 parts as well as partitioning Spare area into 3 parts. That is to have a line along x axis. More, to have files to line up on a single axis.
Raid3,4,5:: is to have files to be placed on layers on a single plane in 3-D.
.
.
Raid6: an ssd or more partitioned into 7 parts as well as the spare area. So we can store files in 3-D space. More, that is to create 3-D storage area.
The rest to come soon. Need to drink a glass of white coloured liquid.
Wear levelling on Raid6 using 2 ssds subdivided identically: Create 7 partitions on each ssd as well as 7 partitions in spare area and stack them together. That will generate 2 cubes of storage area in 3-D space. One to utilise as a storage area and the other to be used as additional spare area. As soon as the valid area and spare area worn out fully, then each file will be sent to next ssd's valid area as their space wear out. Once all the blocks are dead on the first ssd, the files would have been transfered to the next ssd as they were on the initial see. Since the first see is totally worked out, the second can be the main storage area. And we can replace the first see with a new see on a clean state raided6 (that is partitioning valid and spare area with 7 partitions each). And thus we can always have the storage ssd'in fin a state of free of dead blocks. This can be repeated as the storage ssds keep dying on you and have a clean see which is no dead blocks.
Timeout for a glass of white coloured liquid.
Last edited:
Facebook
Twitter