Administering virtual disks

About virtual disks

Virtual disks are used to manage data in a more flexible way on systems with multiple hard disks. They are particularly useful for improving the performance of large applications, such as databases, by distributing the data across multiple disks and speeding up disk I/O.

Units of virtual disk space look like real disk partitions to programs running on the system, but their characteristics can be changed dynamically using the Virtual Disk Manager.

The Virtual Disk Manager adds an additional level of software control to the allocation of data storage. Normally, when applications request some data from the filesystem, the kernel uses the filesystem to discover the disk blocks where the data is stored and returns the data directly. When a virtual disk is in use, the system reads and writes to a virtual disk driver, which in turn manages the physical allocation of data across several disks.

This has a number of advantages. A virtual disk can be assembled from a collection of small disks, or pieces of disks, so that rather than providing a set of small partitions they can be used to provide a single large contiguous disk space. Data can be duplicated (``mirrored'') across drives, so that if one drive succumbs to a hardware failure, the system can continue to operate without interruption: by using a technique called ``striping'', data can be read from and written to disks in parallel, significantly improving I/O performance.

You can select from eight virtual disk types on your SCO OpenServer system; see ``Virtual disk types'' for details. Each provides different disk characteristics, suitable for different types of applications and for different hardware configurations.

See also:

Next topic: Disk arrays and data striping
Previous topic: Administering virtual disks

© 2003 Caldera International, Inc. All rights reserved.
SCO OpenServer Release 5.0.7 -- 11 February 2003