Introduction to RAID controller

1
Click anywhere to start the presentation
2
Introduction to RAID controller
RAID means redundant array of independent disk.
It is a type of storage technology. It combines
multiple disk component into a logical unit. Data
is distributed across several RAID levels.
3
What is RAID
Redundant array of independent disk is a category
of disk drives that has two or more drives in
combination for fault tolerance. These are
frequently used in the server. It allows you to
store data in the multiple places (redundantly)
to improve the storage performance.
4
History of RAID
In the 1987 David Patterson, Garth Gibson and
Randy Katz first published their research paper.
Their paper name was A case of Redundant Arrays
of Inexpensive Disks. Which helps to form the
modern RAID industry. Their paper defined 1 to 5
RAID levels. Today more levels are defined. This
technology remains vibrant as ever.
5
RAID Levels
Number of standard schemes are evolved those are
refer as levels. Originally there ware five
levels but today it vibrantly evolved more
levels. Its level and original data format is
standardize storage networking industry
association (SNIA).
6
Basics of RAID controller
RAID means redundant array of independent disk.
It is a type of storage technology. It combines
multiple disk component into a logical unit. Data
is distributed across several RAID levels.
7
Types of Level
Different architecture are provided by RAID
followed by each architecture provides
performance, capacity and tolerance. These levels
are 0, 1, 2, 3, 4, 5 and 6.
8
Level 0 Striped disk array Without fault
tolerance.
Provides data striping (spreading out files
across multiple disk drives) but without
redundancy. It highly improves the performance.
In certain case if one drive is fail then all
data is lost. Data is fragmented is called
blocks. Number of block is dedicated to the
stripe size. It is configuration parameter of
array. Block is written in the same sector
simultaneously in the different drives.
9
Level 1 Mirroring and Duplexing
Level 1 provides the twice the read transaction
rate in the same disk. Write transaction rate is
also same. Write request update the stripe of
both drives. The array is continue to operate as
long as at one drive is operating. It has
separate controller in each drive for
simultaneously reads called multiplexing.
10
Level 2 Error correcting coding
Strip data at the bit level rather than the block
level. In this level all disk spindle rotating in
a synchronized way that each sequential bit on a
different drive. Across these bits hamming code
is calculated.
11
Level 3 Bit Interleaved Parity
It provides byte level striping with a dedicated
parity disk. It cant service simultaneously it
is used rarely. Parity is calculated and stored
in the dedicated parity drive.
12
Level 4 Dedicated Parity Drive
Commonly used implementation of RAID is block
level striping with a dedicated parity disk. If
disk fails then parity create a replacement.
Disadvantage of it creates bottleneck.
13
Level 5 Block Interleaved Distributed Parity
It provides data striping at the block level. It
performs excellently in the good fault tolerance.
It requires data and all drives but one to be
operate. Array is not destroyed in a single disk
failure. In a single disk failure it reduce the
entire performance. RAID 5 requires at least
three disk.
14
Level 6 Independent data disk with double
parity
It is similar to RAID 5 except instead of one
parity block per stripe. With two independent
parity block RAID 6 can survive loss of two in
among the group of disk.
15
Standardization of RAID
RAID concept is present from in 1986 and it is
installed in the million of configuration. It
lacks of the standard. RAID implementation is
proprietary and unique to every vendor.
16
Software of RAID
RAID is implemented in the server by a software
called JBOD (just a bunch of disks). It acts as a
logical volume manager it manage the parity
calculation.
17
RAID Controller
These are the cards that can be added to a server
and performs the offload of overloaded RAID CPU.
It is a better solution for a single server. For
managing the mirror data it CPU spend no
processing power. It uses either internal drives
or JBOD.
18
Storage Array
It consist of two high performance and redundant
RAID controller. They support multiple RAID
levels and different drive types and speeds.
Storage array also have snap shot, volume copy
etc. There are different types of RAID controller
like IBM RAID controller.
19
Contact Details
188-M Technology Drive Irvine CA 92618 Tel
[phone] Fax [phone] http//www.sho
pricom.com
20
THANK YOU