Dell PowerVault MD3000i Dell PowerVault MD3000/MD3000i Array Tuning Best Pract - Page 31

Dell™ PowerVault MD3000 and MD3000i Array Tuning Best Practices, Definition, RAID 0, RAID 1/10, RAID

Page 31 highlights

Dell™ PowerVault MD3000 and MD3000i Array Tuning Best Practices Term RAID RAID 0 RAID 1/10 RAID 5 RAID 6 RPA Definition directly on the drive in lieu of using an SATA interposer. Redundant Array of Inexpensive Disks RAID Level 0; RAID 0 is a striped set with no redundant information. It is effectively a fully degraded RAID set with no disk redundancy overhead. RAID Level 1/10: The RAID 1/10 implementation on the MD3000/MD3000i follows Berkley RAID 1 standard, expanding it to a redundant N+N mirrored set. Functionally, this implementation is set equivalent to a generic nested RAID 1+0 and operates on as few as 2 physical drives. This allows for any number of drive failures as long as one of each drive pair is available at the cost of half the included physical drives. RAID Level 5; Involves a block-striping algorithm where n-1 disks per stripe in the raid set contain the data and a parity disk P contains parity or checksum used to validate data integrity provide consistency information. The parity is distributed across all drives in a disk group to provided further fault tolerance. RAID 5 provides protection from single drive failure. RAID Level 6; Involves a block-striping algorithm where n-2 disks per stripe in the raid set contain the data and a parity blocks P and Q contains the parity or checksum used to validate data integrity provide consistency information. The parity is distributed across all drives in a disk group to provided further fault tolerance. RAID 6 provides protection from double drive failure. Read Pre-fetch Algorithm: An acronym for the read ahead cache used on the MD3000\MD3000i. RMW RMW2 SAS SATA Saturation SCSI Read, Modify, Write; The second best algorithm available in RAID 5 and RAID 6 write operations. A RMW occurs when a quantity of bits, smaller or equal to an individual segment are modified. A Firmware Generation One adaptation of RMW specifically for Write-Through conditions where write cache is not enabled or requested to a virtual disk. Serial Attached SCSI. Protocol maintained by t10.org Serial Attached Technology Attachment, or Serial ATA. This is the next stage of the legacy Parallel ATA. Primarily in this white paper SATA refers to SATA HDD technology. See Controller Saturation Small Computer System Interface, Protocol maintained December 2008 - Revision A01 Page 31

  • 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

Dell™ PowerVault MD3000 and MD3000i Array Tuning Best Practices
December 2008 – Revision A01
Page 31
Term
Definition
directly on the drive in lieu of using an SATA interposer.
RAID
Redundant Array of Inexpensive Disks
RAID 0
RAID Level 0; RAID 0 is a striped set with no redundant
information.
It is effectively a fully degraded RAID set
with no disk redundancy overhead.
RAID 1/10
RAID Level 1/10: The RAID 1/10 implementation on the
MD3000/MD3000i follows Berkley RAID 1 standard,
expanding it to a redundant N+N mirrored set.
Functionally, this implementation is set equivalent to a
generic nested RAID 1+0 and operates on as few as 2
physical drives.
This allows for any number of drive
failures as long as one of each drive pair is available at
the cost of half the included physical drives.
RAID 5
RAID Level 5; Involves a block-striping algorithm where
n-1 disks per stripe in the raid set contain the data and
a parity disk P contains parity or checksum used to
validate data integrity provide consistency information.
The parity is distributed across all drives in a disk group
to provided further fault tolerance. RAID 5 provides
protection from single drive failure.
RAID 6
RAID Level 6; Involves a block-striping algorithm where
n-2 disks per stripe in the raid set contain the data and
a parity blocks P and Q contains the parity or
checksum used to validate data integrity provide
consistency information. The parity is distributed across
all drives in a disk group to provided further fault
tolerance. RAID 6 provides protection from double drive
failure.
RPA
Read Pre-fetch Algorithm: An acronym for the read
ahead cache used on the MD3000\MD3000i.
RMW
Read, Modify, Write; The second best algorithm
available in RAID 5 and RAID 6 write operations.
A
RMW occurs when a quantity of bits, smaller or equal
to an individual segment are modified.
RMW2
A Firmware Generation One adaptation of RMW
specifically for Write-Through conditions where write
cache is not enabled or requested to a virtual disk.
SAS
Serial Attached SCSI. Protocol maintained by t10.org
SATA
Serial Attached Technology Attachment, or Serial ATA.
This is the next stage of the legacy Parallel ATA.
Primarily in this white paper SATA refers to SATA HDD
technology.
Saturation
See Controller Saturation
SCSI
Small Computer System Interface, Protocol maintained