HP StorageWorks Modular Smart Array 1000 HP Array Configuration Utility User G - Page 67

Selecting a RAID method, Alternative fault-tolerance methods

Page 67 highlights

Item Tolerates simultaneous failure of more than one physical drive RAID 0 No Read performance Write performance Relative cost High High Low RAID 1+0 Only if no two failed drives are in the same mirrored pair High Medium High RAID 5 No High Low Medium RAID 6 (ADG) Yes High Low Medium *Values for the fraction of drive space usable are calculated with these assumptions: (1) all physical drives in the array have the same capacity; (2) online spares are not used; (3) no more than 14 physical drives are used per array for RAID 5; and (4) no more than 56 drives are used with RAID 6 (ADG). Selecting a RAID method NOTE: Not all controllers support RAID 6 (ADG). Most important criterion Fault tolerance Cost effectiveness I/O performance Also important Cost effectiveness I/O performance Fault tolerance I/O performance Cost effectiveness Fault tolerance Suggested RAID level RAID 6 (ADG) RAID 1+0 RAID 6 (ADG) RAID 5 (RAID 0 if fault tolerance is not required) RAID 5 (RAID 0 if fault tolerance is not required) RAID 1+0 Alternative fault-tolerance methods Your operating system may also support software-based RAID or controller duplexing. • Software-based RAID resembles hardware-based RAID, except that the operating system works with logical drives as if they were physical drives. To protect against data loss caused by physical drive failure, each logical drive must be in a different array from the others. • Controller duplexing uses two identical controllers with independent, identical sets of drives containing identical data. In the unlikely event of a controller failure, the remaining controller and drives will service all requests. Neither of these alternative fault-tolerance methods supports online spares or automatic data recovery, nor do they support auto-reliability monitoring or interim data recovery. If you decide to use one of these alternative methods, configure your arrays with RAID 0 for maximum storage capacity and refer to your operating system documentation for further implementation details. Drive arrays and fault-tolerance methods 67

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30
  • 31
  • 32
  • 33
  • 34
  • 35
  • 36
  • 37
  • 38
  • 39
  • 40
  • 41
  • 42
  • 43
  • 44
  • 45
  • 46
  • 47
  • 48
  • 49
  • 50
  • 51
  • 52
  • 53
  • 54
  • 55
  • 56
  • 57
  • 58
  • 59
  • 60
  • 61
  • 62
  • 63
  • 64
  • 65
  • 66
  • 67
  • 68
  • 69
  • 70
  • 71

Drive arrays and fault-tolerance methods
67
Item
RAID 0
RAID 1+0
RAID 5
RAID 6 (ADG)
Tolerates simultaneous failure of
more than one physical drive
No
Only if no two
failed drives are
in the same
mirrored pair
No
Yes
Read performance
High
High
High
High
Write performance
High
Medium
Low
Low
Relative cost
Low
High
Medium
Medium
*Values for the fraction of drive space usable are calculated with these assumptions: (1) all physical drives in the
array have the same capacity; (2) online spares are not used; (3) no more than 14 physical drives are used per array
for RAID 5; and (4) no more than 56 drives are used with RAID 6 (ADG).
Selecting a RAID method
NOTE:
Not all controllers support RAID 6 (ADG).
Most important criterion
Also important
Suggested RAID level
Fault tolerance
Cost effectiveness
I/O performance
RAID 6 (ADG)
RAID 1+0
Cost effectiveness
Fault tolerance
I/O performance
RAID 6 (ADG)
RAID 5 (RAID 0 if fault tolerance is not required)
I/O performance
Cost effectiveness
Fault tolerance
RAID 5 (RAID 0 if fault tolerance is not required)
RAID 1+0
Alternative fault-tolerance methods
Your operating system may also support software-based RAID or controller duplexing.
Software-based RAID
resembles hardware-based RAID, except that the operating system works
with logical drives as if they were physical drives. To protect against data loss caused by physical
drive failure, each logical drive must be in a different array from the others.
Controller duplexing
uses two identical controllers with independent, identical sets of drives
containing identical data. In the unlikely event of a controller failure, the remaining controller and
drives will service all requests.
Neither of these alternative fault-tolerance methods supports online spares or automatic data recovery, nor
do they support auto-reliability monitoring or interim data recovery.
If you decide to use one of these alternative methods, configure your arrays with RAID 0 for maximum
storage capacity and refer to your operating system documentation for further implementation details.