Adaptec MAN-00005-UG, Инструкция по установке и эксплуатации

Страница: 217 / 237

218 Spheras Storage Director Installation and User Guide
An additional measure of fault-tolerance (or improved availability) is achieved using a Hot Spare (Standby) disk. This
disk is powered-on but idle during normal array operation. If a failure occurs on a disk in a fault-tolerant set, the Hot
Spare disk takes over for the failed drive. After completing an automatic rebuild cycle, the array continues to function in
a fully fault-tolerant mode. This means the array can suffer a second drive failure and continue to function before any
disks are replaced.
Impact of Controller Cache on Availability
Every disk array controller has a disk cache. This DRAM or EDRAM physical memory increases the performance of
data retrieval and storage operations. The amount of disk cache varies with the controller model. The controller uses
this memory to store disk writes.
The controller may report to the operating system that the write is complete as soon as the controller receives the data.
This is referred to as Write Back (WB) cache. This improves performance, but exposes you to data loss if a system crash
or power failure occurs before the data in the cache is written to disk.
To avoid potential loss of data, outfit your controller with a battery backup for the cache memory. The battery backup
option enables the cache to be retained until the data can be written to disk. Your data is important. This simple and
inexpensive insurance can prevent a power interruption from turning into a disaster.
Notes: If the array will not use write cache, battery backup is not necessary.
If write cache is used without the battery back-up option, an uninterruptible power supply (also known as a UPS) is
required.
B.5 RAID Levels and Availability
To see the advantages of the RAID levels as they apply to availability, see B.2 RAID Levels on page 214.
B.6 Configuring for Maximum Performance
To see the advantages of the RAID levels as they apply to performance, refer to the following table.
Not Fault
Tolerant No fault-tolerant RAID levels are configured for any of the drives in the
array.
Maximum Performance RAID Level Advantages
RAID Level Access Profile Characteristics
0 Excellent for all types of I/O activity.
1 Excellent for write-intensive applications.
3
Good for sequential or random reads and sequential writes.
5 Excellent for sequential or random reads and sequential writes.
0+1 Excellent for write-intensive applications.
JBOD Mimics normal, individual disk performance characteristics.
Table 5 Maximum performance for maximum performance
Table 4 Maximum availability conditions