Vous êtes sur la page 1sur 2

Individual disk to Storage pool:

Individual Disk ---> JBOD ---> RAID ---> Storage pool

Individual disks cannot address the high storage demand from the production environment.
Day by day Business is increasing customer wanted to develop more applications and want to server
the end user with most effective way.

So To address the High storage demands, Individual disks are bundled together to form a JBOD (Just a
Bunch of Disks).

A JBOD is formed by just grouping the available disks into a bundle.


JBOD just use concatenation technique and form a high capacity disk and later can be presented to the
server.

But, there is no data protection against the disk failures.


In a JBOD of 5 disks, if one disk is faulted, then the data cannot be rebuild and there is no redundancy
or parity among the disks in a JBOD.

JBOD can provide good capacity( without loosing any of the disk capacity for redundancy purpose or
Parity purpose), and JBOD are cost effective and easy management. JBOD have such good features ,
But no protection schemes , hence we cannot risk the valuable data so RAID concept become more
popular.

RAID is also a group of disks, But RAID implemented with fault tolerance against Disk failures and
RAID can also improve performance.

RAID Redundant Array of Inexpensive Disks

RAID uses technique called mirroring for redundancy and other technique called Striping across the
disks for achieving a better IO performance.

Redundancy help to achieve the fault tolerance against disk failures but the effective usable capacity is
only half of the total mirrored RIAD capacity.

So there is another technique introduced , That is Parity concept, parity disk will check for the missed
blocks and help to rebuild the data against the disk failure.

RAID can be configured with particular protection level depending on the performance, capacity, and
redundancy needs of the environment.

So with RAID we overcome the disadvantages of JBOD( data protection and IP performance).

But Still RAID have few limitations and some complications in Storage management using RAID
groups.

In a RAID group we can have only one type of Disks , capacity and rpm values must be same for all
disks under the RAID group.
RAID groups can have a maximum count of 16 drives .
And from RAID group the LUNs will be configured and assigned to the hosts.
As A RAID group can have limited number of disks, Suddenly if we want to increase the Storage
capacity of the RAID group we need to change the configuration settings of RAID.

Suppose we have RAID5(4+1 configuration) , and if we want to increase the capacity of only 300
GB(one disk capacity) but for RAID 5 configuration( 4+1 or 8+1 combination can be possible), so we
need to add 4 more new disks to increase the Storage capacity of existing RAID group and also we
need to change the configuration from RAID(4+1) to RAID(8+1).

So Storage management is not very convenient when we want to add more capacity to the available
RAID groups in Storage array.

So to make the storage management simple, Storage Pool concept is introduced.

Using Storage Pools we can address the limitations of Raid Groups.


Storage pool can be configured with different types of disks, and different capacity disks can also be
used. And there is no limitation of maximum number of disks that can be used inside a Storage pool.
And we can add as many number of disks as per our requirement.

And After implementing Storage pool, we have come up with other idea called Tiering inside pool,
Based on the IO from host, Hot data will be internally moved to Flash storage disks and cold data will
be moved to SATA disks. By doing this we can best utilize the available disks based on their best
feature. And Automatic-Tiering will reduce the burden to Storage admin(No need to plan Storage
layout based on IO patterns), and performance will be increased.

Vous aimerez peut-être aussi