Planning a Fault-Tolerant Disk Configuration

A technology called Redundant Array of Independent Disks (RAID) minimizes loss of data caused by problems with accessing data on a hard disk. RAID is a fault-tolerant disk configuration in which part of the physical storage capacity contains redundant information about data stored on the disks. The redundant information enables regeneration of the data if one of the disks or the access path to it fails, or a sector on the disk cannot be read.

Some vendors sell disk subsystems that implement RAID technology completely within the hardware. Some of these hardware implementations support hot swapping of disks, which enables you to replace a failed disk while the computer is still running Windows NT Server. RAID arrays are not described in this book.

In this chapter, and other chapters in the "Reliability and Recoverability" part of this book, the term RAID array refers to a fault-tolerant disk subsystem where all of the fault tolerance is implemented in the hardware. For information about RAID arrays that are compatible with Windows NT, see the Windows NT Hardware Compatibility List (HCL). Chapter 8, "General Troubleshooting," contains more information about the Windows NT HCL.

Windows NT Server provides software support for two fault-tolerant disk configurations: mirror sets and stripe sets with parity. You use Disk Administrator to configure mirror sets and stripe sets with parity, and to reconstruct the volume when there has been a failure.

Windows NT Server and Windows NT Workstation have software support for stripe sets, which you configure by using Disk Administrator. Although stripe sets are not fault-tolerant, they are described in this section because:

• Stripe sets are the basis for stripe sets with parity.
• Stripe sets improve I/O performance.
• Windows NT Workstation and Windows NT Server need the same kind of information, both on disk and in the Registry, to identify and access stripe sets, mirror sets, and stripe sets with parity.

All RAID disk configurations:

• Regenerate data to satisfy a read request when a disk or a path to a disk has failed. Regeneration involves reading data from other disks in the volume. The regeneration occurs automatically.
• Reconstruct the missing data onto the new disk when you have replaced the failed disk (or the path to it). When you have implemented a fault-tolerant volume by using Disk Administrator, you also have to use Disk Administrator to begin the reconstruction of the data. If you are using a RAID array, see the vendor's documentation for information about reconstructing data.

For complete information about RAID, RAID terminology, and disk arrays, see The RAIDbook, A Source Book for Disk Array Technology. The RAID Advisory Board in St. Peter, MN, publishes this book. The June 1995 edition is the basis for RAID information in this chapter.

Chapter 7, "Protecting Data," in Windows NT Server Concepts and Planning, contains more information about using stripe sets, mirror sets, and stripe sets with parity on Windows NT Server.

Chapter 5, "Preparing for and Performing Recovery," in this book, describes configuring, and recovering from failures of, mirror sets and stripe sets with parity.