Administrator Guide

Best practices
This appendix describes best practices for configuring and provisioning a storage system.
Topics:
• Pool setup
• RAID selection
• Disk count per RAID level
• Disk groups in a pool
• Tier setup
• Multipath configuration
• Physical port selection
Pool setup
In a storage system with two controller modules, try to balance the workload of the controllers. Each controller can own one virtual pool.
Having the same number of disk groups and volumes in each pool will help balance the workload, increasing performance.
RAID selection
A pool is created by adding disk groups to it. Disk groups are based on RAID technology.
The following table describes the characteristics and use cases of each RAID level:
RAID level
Protection Performance Capacity Application use
cases
Suggested disk
speed
RAID 1/RAID 10
Protects against up
to one disk failure
per mirror set
Great random I/O
performance
Poor: 50% fault
tolerance capacity
loss
Databases, OLTP,
Exchange Server
10K, 15K, 7K
RAID 5 Protects against up
to one disk failure
per RAID set
Good sequential I/O
performance,
moderate random
I/O performance
Great: One-disk fault
tolerance capacity
loss
Big data, media and
entertainment
(ingest, broadcast,
and past production)
10K, 15K, lower
capacity 7K
RAID 6 Protects against up
to two disk failures
per RAID set
Moderate sequential
I/O performance,
poor random I/O
performance
Moderate: Twodisk
fault tolerance
capacity loss
Archive, parallel
distributed file
system
High capacity 7K
Disk count per RAID level
The controller breaks virtual volumes into 4-MB pages, which are referenced paged tables in memory. The 4-MB page is a fixed unit of
allocation. Therefore, 4-MB units of data are pushed to a disk group. A write performance penalty is introduced in RAID-5 or RAID-6 disk
groups when the stripe size of the disk group isn't a multiple of the 4-MB page.
• Example 1: Consider a RAID-5 disk group with five disks. The equivalent of four disks provide usable capacity, and the equivalent of
one disk is used for parity. Parity is distributed among disks. The four disks providing usable capacity are the data disks and the one
disk providing parity is the parity disk. In reality, the parity is distributed among all the disks, but conceiving of it in this way helps with
the example.
Note that the number of data disks is a power of two (2, 4, and 8). The controller will use a 512-KB stripe unit size when the data disks
are a power of two. This results in a 4-MB page being evenly distributed across two stripes. This is ideal for performance.
C
166 Best practices