Partitioning a hard disk is the act of defining areas of the disk for an operating system to use as a volume. When you partition a disk, the partitioning software writes a master partition boot sector at Cylinder 0, Head 0, Sector 1—the first sector on the hard disk.

This sector contains data that describes the partitions by their starting and ending cylinder, head, and sector locations. The partition table also indicates to the ROM BIOS which of the partitions is bootable and, therefore, where to look for an operating system to load.

The FDISK program is the accepted standard for partitioning hard disk drives for use with all operating systems up through Windows Me. Windows 2000 and XP use a similar command-line program called DISKPART, or you can partition and format hard disks with the Disk Management tool.

All versions of Windows starting with Windows 95 can also partition and format the drive using the SETUP program for installing the OS. These programs are included with your operating system, and although they might have the same name and basic functions with any OS, you should typically use the tools that specifically came with your OS.

Partitioning prepares the boot sector of the disk in such a way that the FORMAT program or the Windows SETUP format utility can operate correctly. Partitioning also enables various operating systems to coexist on a single hard disk. No matter which operating system you use, it should come with an FDISK, DISKPART, or SETUP program that can be used to partition the drive.

With any version of Windows, as with MS-DOS, FDISK or other disk preparation tools enable you to create two types of disk partitions: primary and extended. A primary partition can be bootable but an extended partition can't. If you have only a single hard disk in your system, at least part of the drive must be prepared as a primary partition if you want to start your computer from the hard disk (and who doesn't?).

A primary partition is seen as a single volume or drive letter (C: on one-drive systems), whereas an extended partition acts as a sort of logical container for additional volumes (drive letters D: and beyond). A single extended partition can contain a single volume (also referred to by FDISK as a logical DOS drive) or several volumes (logical DOS drives) of various sizes.

Don't get hung up on the fact that FDISK calls partitions "DOS" partitions or "DOS" drives. This is true even though the operating system you are installing is Windows 95, 98, Me, NT, 2000, XP, Linux, and so on. Depending on the version of Windows in use (and with any version of MS-DOS), you might need to subdivide a hard drive through the use of FDISK.

The original release of Windows 95 and all MS-DOS versions support a file system known as FAT16, which allows no more than 65,536 files per drive and a single drive letter or volume size of no more than 2.1GB. Thus, a 10GB hard disk prepared with MS-DOS or the original Windows 95 (or 95A) must have a minimum of five drive letters and could have more.

Another reason for subdividing a single drive into multiple volumes is increased data security. For example, PowerQuest (the creator of PartitionMagic) suggests a three-volume partitioning scheme that looks like this:

C: for the operating system and utilities

D: for applications

E: and above for data

In this example, primary and extended partitions would be assigned as shown here:

C: the primary partition

D: and E: in the extended partition as logical DOS drives (volumes)

If a catastrophic failure to the disk structure wipes out C: or D:, drive E:—which contains the data—is usually still intact. This method also makes backing up the data easier; just set the backup program to back up all of E:, or all the changed files on E:.

Large Hard Disk Support

If you use the Windows 95B or above (Win95 OSR 2.x), Windows 98, Windows Me, or Windows 2000 versions of FDISK with a hard drive greater than 512MB, FDISK offers to enable large disk support.

Choosing to enable large disk support provides several benefits:

• You can use a large hard disk (greater than 2.1GB) as a single drive letter. In fact, your drive can be as large as 2TB and still be identified by a single drive letter. This is because of the FAT32 file system, which allows for many more files per drive than FAT16.

• Because of the more efficient storage methods of FAT-32, your files will use less hard disk space overall.

However, keep in mind that all disk operations must be performed through an operating system with FAT32 support (Windows 95B or later, Windows 98, Windows Me, or Windows 2000). If you have old MS-DOS games or applications that are bootable, you won't be able to access a FAT32 volume unless you replace the DOS version on the disk with the DOS included on Win95B or Windows 98.

This normally can be done by using the SYS A: command from the \Windows\Command folder. Another option is to use the Windows Startup menu in Win95B or Win98 (press F8 as Windows starts to load) and select Command Prompt to get to a FAT32-capable DOS. On the other hand, you can select Start, Shutdown, Restart the Computer in MS-DOS Mode from the Windows desktop.

A drive prepared with the large hard disk support (FAT32) option enabled can still be partitioned into primary and secondary partitions with FDISK, as with FAT16 for the data-security reasons listed earlier. Another type of file system is NTFS, which is supported by only Windows NT, Windows 2000, and Windows XP.

This is a high-performance file system with additional security and networking features. NTFS was revised to NTFS5 in Windows NT 4.0 with Service Pack 4 and later, as well as Windows 2000 and Windows XP. Note that Windows 9x can't read NTFS partitions, and Windows NT can't read FAT32 partitions. Windows 2000 and Windows XP, however, can handle both FAT32 and NTFS.

Assigning Drive Letters with FDISK

FDISK can be used in many ways, depending on the number of hard drives you have in your system and the number of drive letters you want to create. With a single drive, creating a primary partition (C:) and an extended partition with two logical DOS drives within it results in the following drive letters:

Partition Type	Drive Letter(s)
Primary	C:
Extended	D: and E:

Most people think that a second physical drive added to this system should have drive letters that follow the E: drive. However, you must understand how drive letters are allocated by the system to know how to use FDISK correctly in this situation. Table below shows how FDISK assigns drive letters by drive and partition type:

Drive	Partition	Order	Drive Letter
1st	Primary	1st	C:
1st	Extended	3rd	E:
2nd	Primary	2nd	D:
2nd	Extended	4th	F:

How does this affect you when you add another hard drive? If you prepare the second hard drive with a primary partition and your first hard drive has an extended partition on it, the second hard drive takes the primary partition's D: drive letter. This moves all the drive letters in the first hard drive's extended partition up one drive letter.

In the first example, a drive is listed with C: (primary partition), D:, and E: (extended partition volumes) as the drive letters (D: and E: were in the extended partition). Table below indicates what happens if a second drive is added with a primary partition and an extended partition with two volumes (same setup as the first drive).

Drive	Partition Type	Order	Original Drive Letters (First Drive Only)	New Drive Letters After Adding Second Drive
1st	Primary	1st	C:	C:
1st	Extended	3rd	D:, E:	E:, F:
2nd	Primary	2nd		D:
2nd	Extended	4th		G:, H:

After adding the second drive, the original drive letters D: and E: have now become E: and F:. The primary partition on the new drive has become D:, and the extended partition volumes on the second drive are G: and H:. Confused? Well, you better not be or you'll find yourself deleting or copying data to or from the wrong drive.

This principle extends to third and fourth physical drives as well: The primary partitions on each drive get their drive letters first, followed by logical DOS drives in the extended partitions.

One way to effect this is to partition additional drives with only extended partitions—in other words, do not create primary partitions on them. That enables the new drive's partitions to be seen only as additional letters, and the letters used by the first drive's partitions to remain unchanged.

If you're adding a drive to your system, you should now understand why preparing that second, third, or fourth drive with a primary partition is a bad idea. If you're installing an additional hard drive (not a replacement), remember that it can't be a bootable drive.

And if it can't be bootable, there's no reason to make it a primary partition. FDISK allows you to create an extended partition using 100% of the space on any drive. Table below shows the same example used in Table above with the second drive installed as an extended partition.

Drive	Partition Type	Order	Original Drive Letters (First Drive Only)	New Drive Letters After Adding Second Drive
1st	Primary	1st	C:	C:
1st	Extended	3rd	D:, E:	D:, E:
2nd	Primary	2nd	—	—
2nd	Extended	4th		F:

When a new drive is added with only extended partition volumes, you can see that the original drive letters remain undisturbed.

This arrangement is much easier to understand, and it prevents accidents with data because of drive letters changing. This operating system behavior also explains why some of the first computers with ATA-based (ATAPI) Iomega Zip drives had the Zip drive as D:, with a single hard disk identified as C: and E:.

The Zip drive was treated as the second hard drive with a primary partition. Subsequently, Iomega changed the Zip drive format and driver so that the Zip disk is recognized as an extended partition at the end of the drive letter chain.