1. Presented By : Ishwor Panta
Roll No. : 13
Presentation on RAID
2. RAID
Redundant Array of Independent Disks
Example of storage virtualization
Combines multiple disk drives into logical unit
Provides data integrity, fault tolerance, throughput
compared to single driver
Typically used in server computer and advanced
personal computers
3. RAID
Basic Idea is to connect multiple disks to provide :
Large Storage Capacity
Faster Access to reading data
Redundant Data
Increase Performance
4. Working of RAID
Combine multiple independent inexpensive disk
and group them together
OS sees the combined disk as a single logical disk
RAID Controller Card is used to handle all I/O to
disks and locate the stored data
RAID Controller is a hardware device or software
program used to manage array of Storage devices.
5. RAID TERMINOLOGY
Mirroring :
Copying of data to multiple hard disks
Striping:
Splitting of data across multiple hard disks
Error Correction:
Detecting hard disk failure and recovering from
the failure, done with parity checking
7. RAID 0
Striping
Offer superior I/O
Performance can be enhanced
by using multiple controller
Requires minimum of 2 disks
No fault tolerance
8. RAID 1
Mirroring
Controller uses either the
data drive or the mirror drive
for data recovery
Two minimum disk required
Storage capacity is half of
total disk capacity
9. RAID 2
Redundancy through
hamming code
Stripe data at bit level
rather than block
Minimum 3 disk are required
Data transfer rate is very high
Single bit error can be detected and corrected
10. RAID 3
Bit interleaved parity
Files are broken into
“byte-sized pieces”
Written in parallel in two
or more drives
Additional drive stores parity information
High throughput for large data transfer
Minimum of 3 disk are required
11. RAID 4
Block interleaved parity
High read rate but slow
write rate
Each file is split into blocks
of certain size
Service multiple read request simultaneously
Fault tolerance is very high
Three minimum disk are required
12. RAID 5
Striping with parity
Most secure RAID
Parity information is
spread across all drives
Read data transactions are
very fast
Disk failure may have an effect on throughput
Complex technology
13. RAID 6
Extension of RAID 5 level
with additional fault
tolerance (independent
parity scheme)
Data is striped as like in
RAID 5 on block level
Can sustain from multiple drive failures
Poor write speed and complex system
14. RAID 1+0
Uses multiple RAID1
in single array
Data striped across
all mirrored sets
High fault tolerance
High performance rate
High implementation cost