User manual

Chapter 1 RAID Introduction
1.1. What is RAID?
It is inevitable that a single hard disk, either with SCSI or IDE interface, will suffer the compatibility
problem between the motor rotational speed and the transfer interface. As a result, an Ultra160 SCSI or
ATA100 IDE hard disk can only achieve the transfer rate of up to 30MB/Sec at the bandwidth of 100MHz.
In addition, the life span of a hard disk is limited. Once a hard disk is damaged, it is likely to lead to the
system crash and data loss. Abovementioned are two severe problems in network system architecture.
These factors have encouraged the generation of RAID (Redundant Arrays of Inexpensive / Independent
Disks), a technology that combines multiple inexpensive and independent hard disks into an array of hard
disks so as to increase data transfer performance and storage effi ciency.
RAID implements the mechanisms such as Striping or Mirroring plus Parity Checking, to combine two
or more physical hard disks into one virtual/logical disk array that allows On-line, quick access, huge
capacity and fault tolerance. If one of the hard disks in the array is damaged, the system continues to
operate using the remaining working hard disks, resulting in no system crash or data loss.
In summary, RAID technology increases the performance and data security in network system architecture.
1.2. RAID Functions
• Expanding storage capacity
• Increasing data transfer speed
• Saving cost
• Inherent Fault Tolerance
• Hot Swap
• Auto-Rebuild
• Hot Spare
• On-line Capacity Expansion
1.3. RAID Levels and Comparisons
With the development of RAID technologies, users now have more options for RAID levels. Some
vendors even have proprietary names for RAID levels. Generally, RAID levels include RAID 0, 1, 3, 5, 6,
0+1, 10, 30, 50, 60 and JBOD (Just a Bunch Of Disks).
The following is a comparison of some commonly used RAID levels:
RAID Level
Basic
Operation Method
Hard Disk Available Capacity Data Reliability
Data
Transfer Speed
Minimum
Number of
Hard Disks
0
Striping data across each
drive
Total capacity of all the hard
disks
Low Highest 2
1 Mirroring
Half of total capacity of all the
hard disks
High Lower 2
0+1
Mirroring data to another
drive
and stripping across the
drive
Half of total capacity of all the
hard disks
Very High High 4
3
Stores parity information on
independent disk
Total capacity of all the hard
disks minus one disk capacity
High Very High 3
5
Stores parity information on
all the hard disks
Total capacity of all the hard
disks minus one disk capacity
High Very High 3
3+Spare
Stores parity information
on independent hard disk &
spare disk
Total capacity of all the hard
disks minus two disks capacity
Very High High 4
5+Spare
Stores parity information on
all the hard disks & spare
disk.
Total capacity of all the hard
disks minus two disks capacity
Very High High 4
For more information of RAID, refer to our website at www.stardom.com.tw
1.4. Storage Introduction
A connection using storage protocol such as SCSI requires “two end points”. These two end points are
called “initiator” and “target”. SCSI initiator will request for the operation of all SCSI, such as reading and
writing. It is usually on the host or server side. (such as SCSI HBA)
On the other hand, target refers to the storage equipments or the devices used for managing or allocating
capacity. Basically, it is a device for executing SCSI commands. A target can be a disk, tape or disk array.
1.5. Glossary
The following terms will be used throughout this manual:
RAID Redundant Array of Independent Disks DS Dedicated Spare disks
PD Physical Disk GS Global Spare disks
VG Volume Group DC Dedicated Cache
UDV User Data Volume GC Global Cache
CV Cache Volume DG DeGrade mode
LUN Logical Unit Number S.M.A.R.T. Self-Monitoring Analysis and Reporting
Technology.
GUI Graphic User Interface. WWN World Wide Name.
WT Write-Through HBA Host Bus Adapter.
WB Write-Back MPIO Multi-Path Input/Output.
RO Read-Only SAF-TE SCSI Accessed Fault-Tolerant Enclosures.
18
RAID Introduction RAID Introduction 19