Table of Contents
Advertisement
5.2.3 Data Parity
The controller can generate "parity" for the ability to protect and rebuild data. Parity protects
stored information without requiring data mirroring. When data is protected by parity, it is still
available if a drive fails. Parity-protected data is reconstructed using the parity formula. You can
remove and replace a failed disk (known as "hot swapping"), and the controller then rebuilds the
data using the information on the remaining drives.
5.3 RAID Levels
The ExaSAN RAID system supports several RAID levels and configurations. Each level has a
different architecture and provides varying degrees of performance and fault tolerance. Each
level has characteristics to achieve maximum performance or redundancy depending on the
data environment.
5.3.1 RAID 0: Striping
RAID level 0, striping only, is the fastest and most efficient array type, but offer no fault-
tolerance. Any drive failure destroys the data in the array.
5.3.2 RAID 1: Mirroring
RAID level 1, mirroring, has been used for Metadata LUN because of its simplicity and high
levels of reliability and availability. Mirroring uses two drives, each drive stores identical data.
RAID 1 provides very high data reliability and improved performance for read-intensive
applications, but this level has a high capacity cost because it retains a full copy of your data on
each drive in mirror set.
In a RAID 1 configuration, the capacity of the smallest drive is the maximum storage area.
5.3.3 RAID 5: Independent data disks with distributed parity
By distributing the parity information across all drives in a set, RAID level 5 achieves high
reliability and data availability. It also offers the highest read data transaction rate of all levels
along with a medium write rate. The low ratio of ECC (Error Correction Code) parity disks to
data disks offers hardware efficiency. Disk failure has a moderate impact on the total transfer
rate.
5.3.4 RAID 6: Independent data disks with two Independent parity schemes
RAID level 6 extends RAID level 5 by adding an additional parity block; thus it uses block-level
striping with two parity blocks distributed across all member disks. RAID 6 does not have a
performance penalty for read operations, but it does have a performance penalty on write
operations because of the overhead associated with parity calculations.
RAID 6 is no more space inefficient than RAID 5 with a hot spare drive when used with a small
number of drives, but as arrays become bigger and have more drives the loss in storage
capacity becomes less important and the probability of data loss is greater. RAID 6 provides
protection against data loss during an array rebuild, when a second drive is lost, a bad block
read is encountered, or when a human operator accidentally removes and replaces the wrong
disk drive when attempting to replace a failed drive.
User Guide
RAID
5-2
Advertisement
Table of Contents
