Databases Reference
In-Depth Information
depending on the workload. One thing to watch out for is RAID controllers that
use a “concatenated mirror” implementation for RAID 10. This is suboptimal be-
cause of the absence of striping: your most frequently accessed data might be placed
on only one pair of spindles, instead of being spread across many, so you'll get
poor performance.
RAID 50
RAID 50 consists of RAID 5 arrays that are striped, and it can be a good compro-
mise between the economy of RAID 5 and the performance of RAID 10, if you have
many disks. This is mainly useful for very large datasets, such as data warehouses
or extremely large OLTP systems.
Table 9-2
summarizes various RAID configurations.
Table 9-2. Comparison of RAID levels
Level
Synopsis
Redundancy
Disks required
Faster reads
Faster writes
RAID 0
Cheap, fast, dangerous
No
N
Yes
Yes
RAID 1
Fast reads, simple, safe
Yes
2 (usually)
Yes
No
RAID 5
A safety, speed, and cost compromise cost
Yes
N + 1
Yes
Depends
RAID 10
Expensive, fast, safe
Yes
2N
Yes
Yes
RAID 50
For very large data stores
Yes
2(N + 1)
Yes
Yes
RAID Failure, Recovery, and Monitoring
RAID configurations (with the exception of RAID 0) offer redundancy. This is impor-
tant, but it's easy to underestimate the likelihood of concurrent disk failures. You
shouldn't think of RAID as a strong guarantee of data safety.
RAID doesn't eliminate—or even reduce—the need for backups. When there is a
problem, the recovery time will depend on your controller, the RAID level, the array
size, the disk speed, and whether you need to keep the server online while you rebuild
the array.
There is a chance of disks failing at exactly the same time. For example, a power spike
or overheating can easily kill two or more disks. What's more common, however, is
two disk failures happening close together. Many such issues can go unnoticed. A
common case is corruption on the physical media holding data that's seldom accessed.
This might go undetected for months, until either you try to read the data, or another
drive fails and the RAID controller tries to use the corrupted data to rebuild the array.
The larger the hard drive is, the more likely this is.
That's why it's important to monitor your RAID arrays. Most controllers offer some
software to report on the array's status, and you need to keep track of this because you
might otherwise be totally ignorant of a drive failure. You might miss your opportunity
to recover the data and discover the problem only when a second drive fails, when it's
