Databases Reference
In-Depth Information
To understand a RAID 0 set, imagine a series of four disk drives lined up in a row. Data written to
a stripe set will i ll the i rst drive with a small amount of data. Each subsequent drive will then be
i lled with the same amount of data, at which point the process is repeated starting with the i rst
disk drive. Figure 4-3 shows how data looks after it has been written to a RAID 0 disk subsystem.
Each data stripe is made up of some uniform data size. Most RAID systems allow the user to mod-
ify the size of the data stripe.
Data 1
Data 2
Data 3
Data 4
Data 5
Data 6
Data 7
Data 8
Data 9
Data 10
Data 11
Data 12
Data 13
Data 14
Data 15
Data 16
FIGURE 4-3
Concatenated disk arrays are similar to stripe datasets, differing in the method used to load data.
You can think of concatenated datasets as a group of disk drives that are i lled in series. The i rst
group is i lled, then the second group, and so on. We will investigate the performance implications
of different RAID coni gurations later in the Disk Drive Performance section of this chapter.
Figure 4-4 shows the contrast between striped RAID, which is serpentine in its layout, and the
waterfall pattern of a concatenated disk array. Concatenated systems don't necessarily lack data pro-
tection. Many storage arrays layer different types of RAID. One example is a system that combines
mirrored physical disks into a concatenated RAID set. This combined system offers the benei ts of
protected data and the ease of adding more capacity on demand since each new concatenated mirror
will be appended to the end of the overall RAID set.
Data 1
Data 5
Data 9
Data 13
Data 2
Data 6
Data 10
Data 14
Data 3
Data 7
Data 11
Data 15
Data 4
Data 8
Data 12
Data 16
FIGURE 4-4
Search WWH ::
Custom Search