HARD DRIVE TEXT

Zip drives

A zip drive is similar to a floppy drive, but it is a little bit larger physically and stores more information. A zip drive can store 100MB or 250MB of data, depending on what type of drive it is. These drives were popular solutions for backing up data before CD-ROMs and external drives became popular.

Cylinders

All the platters in the hard disk contain the same number of tracks, but that number varies from one hard disk to another. These tracks are numbered from the outside in, starting with zero. For example, on a platter with 10 tracks, the track closest to the outer edge of the platter is Track 0, while the track closest to the center is Track 9. A cylinder consists of the same track on both sides of all the platters. In other words, when you reference Track 0, you reference a particular track on a particular platter, but when you reference Cylinder 0, you reference Track 0 on all platters. If you know the number of cylinders, heads, and sectors per track, you can calculate the size of a disk. For example, if a drive has 4,092 cylinders, 16 heads, and 63 sectors per track, the size of the disk would be 2,111,864,832 bytes (2.1GB). The formula to calculate the size of the disk is number of cylinders × number of heads × number of sectors per track × 512 bytes per sector.

lba important point

In order to leverage larger size drives, your BIOS would have to support LBA or ECHS — which most BIOS do today. Notice that an LBA enabled BIOS can only support an 8.4GB drive — and we are way past that drive size today. Today's BIOS support the INT13 extensions, developed by Phoenix Technologies, which allow the systems to see drives past 137GB in size! The BIOS can recognize larger size drives because it simply identifies the drives by the number of sectors.

reasons to dual boot

Reasons to dual boot There are a number of reasons you may want to have a dual-boot system: Software doesn't run smoothly on one operating system, so you install an additional operating system and boot to that OS to run your software. You are part of the help desk team in an organization that runs two different operating systems, and you will need to boot to the appropriate OS to find solutions to problems. You are studying for the A+ exam and need to run Windows 2000 and Windows XP but don't have two different computers.

Performance

Seek time is the time it takes to move the read/write heads to the desired track. Seek time is a calculated average because the time it takes to move to the desired track differs from one instance to another. For example, if the read/write heads are on Track 1, they will take longer to move to Track 12 than to Track 3. Seek time is measured in milliseconds, or one thousandth of a second. Figure 5-4 shows how seek time is measured. ✦ Latency is the time it takes for the appropriate sector to move under the read/write head. Latency is measured in milliseconds. ✦ Access time describes the overall speed of the disk. It is a combination of seek time and latency. The lower the access time, the better. ✦ Spin speed is the speed at which the platters spin, measured in rotations per minute, or rpm. The larger the rpm value, the faster the disk, which means less latency.

Primary partition

This is the partition that the computer boots from; the operating system's boot files are loaded from here. You are allowed to have four primary partitions per disk. Because you may have multiple primary partitions (say, if you're running several operating systems on the same computer), you must designate one primary partition as the active partition — the partition from which your normal operating system loads.

Understanding Management ToolsAfter partitioning the disk and formatting the drive, some maintenance still needs to be done on a regular basis. Drive maintenance helps you address two areas of concern that have an impact on a user's data: ✦ File system optimization: This means getting the best performance from your file system. If you can open the files stored on a computer but they open much more slowly than they did two months ago, your system needs optimizing. Figure 5-41: Viewing the RAID 5 volume in disk management. Figure 5-40: Select the disks to be members of the RAID 5 volume and specify the size of the volume.

After partitioning the disk and formatting the drive, some maintenance still needs to be done on a regular basis. Drive maintenance helps you address two areas of concern that have an impact on a user's data: ✦ File system optimization: This means getting the best performance from your file system. If you can open the files stored on a computer but they open much more slowly than they did two months ago, your system needs optimizing. Figure 5-41: Viewing the RAID 5 volume in disk management. Figure 5-40: Select the disks to be members of the RAID 5 volume and specify the size of the volume.

Flash cards

Multimedia devices, such as MP3 players and digital cameras, now support adding memory to the device by using flash cards (shown in Figure 5-20). The benefit of flash cards is that most computers and laptops have ports on them that you can insert the memory card into. This allows you to upload pictures or music to the memory without needing to connect the camera, which requires connecting a USB cable to the computer.

EIDE/ATA-2 standard

The Enhanced Integrated Drive Electronics (EIDE) standard followed shortly after the IDE standard. The EIDE standard allows four drives to be connected to a dual-channel controller. This is usually implemented as a motherboard with two connectors, also known as controllers, one primary and one secondary. You could then connect two drives off of each controller, making a master/ slave chain for each controller. Figure 5-5 shows a primary controller and a secondary controller on the motherboard. EIDE also supports larger hard disks than the original IDE standard; the original size of an IDE drive was approximately 528MB. The EIDE standard is also known as ATA-2 standard. Most techies in the computing industry interchange the terms IDE and EIDE. For example, if you have a newer motherboard that uses EIDE drives, most documentation and technicians simply use the term IDE — although the drives may be EIDE, or even an Ultra DMA drive. Also note that most techies and documentation use the term controller for the connectors on the motherboard that you connect the drives to — but really the controller is the circuitry on the hard drive that controls access to the drive. My point is get used to people using the terms IDE instead of EIDE and controller instead of connectors.

CD-ROM/CDRW

A Compact Disc-Read Only Memory (CD-ROM) is an optical storage technology that uses a laser to read and write data. Originally, as the name implies, one could only read from CD-ROMs. CD-ROMs are the preferred media to distribute software and fairly large amounts of data. Originally, CD-ROMs stored 650MB of data and could store 74 minutes of music, but today's CD-ROMs store 700MB of data or 80 minutes of music. CD-ROMs are written to from the inside out, and if you look at the bottom of the CD-ROM, you will see the lines indicating where data was written — very useful information if you ever pick up a CD and wonder if it was written to. The speed (transfer rate) of the CD-ROM is measured in multiples of 150 KBps and is indicated with an X. For example, an old 1X (pronounced "one times") CD-ROM had a transfer rate of 150 KBps, while an 8X has a transfer rate of 1200 KBps (150 × 8 = 1200), and a much faster 48X has a transfer rate of 7200 KBps. Today, CD drives are writeable: if you want to burn your own CD you can. (Writing to a CD is often called burning a CD.) This makes CD drives much more popular than the older floppy drives due to the amount of information you can store on the CD — a great solution for backing up your data! There are two types of writeable CDs, a CD Recordable (CDR) and a CD ReWritable (CDRW). The difference between a CDR and CDRW is that youcan write to the CDR only once, while you can reuse the CDRWs many times by reformatting the CD and starting again. A CDR is an example of a Write Once Read Many (WORM) disc. You can write to a CDR multiple times, but you cannot overwrite areas of the disc that have already been written to. This means that with a CDR if you write to it many times the additional write operations are appending the information to the end of the CDR. Each burn operation that you perform is called a session, and most writeable CD drives today support multiple sessions. For example, say you back up your pictures to CD and use only 350MB Book II Chapter 5 Working with Storage Using Removable Storage 235 of space on the CDR. You can write more data to the CD with the remaining space at a later time. With the cost of CDR being so low nowadays, I typically don't bother. I burn a CD, label it, and then file it away. When writing to a CD, the process is not done magnetically like it is with hard drives — as mentioned earlier, the write operation is performed with a laser. The CD has a chemical-dye layer mixed with a thin reflective layer. When you write to the CD, the chemical layer is heated with the laser to create an solid state at that location. These locations reflect less light, and the different patterns of reflection create the data on the disk.

Dynamic disks

A dynamic disk doesn't use partitions but rather volumes as discrete units of space. Because you are creating volumes instead of partitions, you don't have the four-partition limitation of basic disks. With dynamic disks, you are allowed to create as many volumes as you wish. When creating a volume on a dynamic disk, you can create a number of different volume types: ✦ Simple volume: A simple volume is just a block of space that is similar in concept to a partition. ✦ Striped volume: A striped volume is a volume that is made up of equal space across multiple hard disks. With striped volumes, when you save a file to the volume, the file is saved across both disks at the same time. The benefit of a striped volume is a performance benefit from the fact that multiple disks are working at the same time to save the file. ✦ Spanned volume: A spanned volume is a volume that is made up of unequal amounts of space that span multiple disks. The benefit of spanned volumes is that you can join multiple areas of free space to create a single volume that users can access through a single drive letter. ✦ Mirrored volume: A mirrored volume is a volume that is made up of two disks. Data that is written to the volume is stored on both disks, each with a full copy of the data. If one of the disks should fail, the other disk has a copy of the data. Mirrored volumes are only supported on the server versions of the Windows operating systems. ✦ RAID 5 Volume: A RAID 5 volume uses between 3 and 32 disks. Data saved to the volume is spread across all disks in the volume, along with parity data. The parity data is used to calculate data that is unreadable due to a failed disk in the volume. RAID 5 volumes are also available only on server versions of the Windows operating system, so you won't be able to create one on Windows XP. Creating a volume on a dynamic disk in Windows 2000/XP/2003 requires you to first upgrade the basic disk to a dynamic disk. After you upgrade the disk to a dynamic disk, you will notice that the create partition command has changed to a create volume command in the Disk Management console. To upgrade the basic disk to a dynamic disk, follow these steps: 1. Click Start, right-click My Computer, and choose Manage. 2. When the Computer Management console has started, select Disk Management on the left side of the screen. 248 Managing Partitions and Volumes On the right side, in the bottom half of the screen, notice that the disk type for Disk 1 is Basic Disk. 3. Right-click the disk and choose Convert to Dynamic Disk, as shown in Figure 5-26.4. In the Convert to Dynamic Disk dialog box, ensure that the disk number you wish to convert is selected and click OK. 5. Click the Convert button on the Disks to Convert screen, as shown in Figure 5-27.You get a warning letting you know that you will be unable to start older operating systems from the disk after it is converted. 6. Click Yes in the Warning dialog box. 7. Click Yes to dismount any file systems being converted. After the drive is converted, you will notice that the legend has changed to include a simple volume. You will also notice that the disk is now a dynamic disk. Now that you have converted a basic disk to a dynamic disk, you are ready to create a volume in Windows XP. Creating a volume is similar to creating a partition. To create a volume, follow these steps:1. In the Disk Management console, right-click Unallocated Space and choose New Volume, as shown in Figure 5-28.

Clusters

A group of sectors makes up a cluster, and a cluster is the allocation unit for a file — meaning where a file is saved. When a partition (a partition is a logical division of space on the disk; refer to the section titled "Managing Partitions and Volumes," in this chapter) is formatted, the file system determines the cluster size based on the partition size. For example, a 2GB FAT partition uses a 32K cluster size. That same 2GB partition formatted as FAT32 uses only a 4K cluster size. Having a partition use a 4K cluster size means that eight sectors make up a cluster. Keep in mind that after a file has been saved to the cluster, no other file can occupy that cluster. For example, if you have a 32K cluster size and you save a 3K file to the hard disk, the file is saved to an empty cluster — but only 3K of that cluster is used, and the remaining 29K is empty. The remaining 29K is now considered unusable space; no other file can be saved to that unused 29K.

Tape

A popular type of media for data is tape, which is typically used to store backup copies of the data. This copy of the data stored on tape is used to bring the data back if the hard drive fails. Different types of tape drives are popular today: Figure 5-20: A Compact- Flash card used to store data on a digital camera. Figure 5-19: A flash drive uses a USB connection to interface with the computer. Book II Chapter 5 Working with Storage Understanding File Systems 237 ✦ Quarter-Inch Cartridge (QIC): QIC is one of the oldest and most popular tape standards that supports many different tape sizes. The QIC size depends on the standard, and each standard is labeled similar to QIC-80 — the QIC tape that supports 80MB of data. Other examples are QIC-40 (40MB tapes) and QIC-5210 (25GB tapes). ✦ Travan: Travan is a tape drive standard that is based on the QIC standard but supports compression. Travan is typically more expensive and can store from 400MB to 8GB of data or more. Examples of tape sizes in this standard are labeled as TR-1, which stores 400MB of data, and TR-4, which stores 8GB of data. ✦ Digital Audio Tape (DAT): DAT drives store data in a digital format and use two heads, one for reading data and the other for writing data. DAT drives use a standard called Digital Data Storage (DDS) to store data. Examples of DDS tapes are DDS-1, which can store 2GB of data, DDS-2, which can store 8GB of data, and DDS-4, which can store 40GB of data.

Defragmentation utility

After a drive is formatted and you start storing information on it, the information is written to one cluster at a time. This means that on a freshly formatted drive, the contents of a file are written to clusters on the disk that are side by side. This ensures optimal performance when opening a file because the read/write heads don't have to jump from one end of the disk to another to open a single file. Unfortunately, the disk won't stay in this state because as you add to and delete files, the contents of these files are scattered throughout the disk. The following minitable shows a fragmented disk. In this example, you can see that block 2, 5, and 7 belong to the same file but are scattered throughout the disk. This causes a performance decrease when accessing the file because the read/write heads need to locate the contents of the file that are spread throughout the disk. Disk defragmenting applications clean this up by taking all the data of a single file and placing it in clusters that reside side by side. You can see below what the disk would look like after a Windows defragmentation.Windows 2000/XP/2003 has a Disk Defragmenter utility in the Computer Management console. This great little tool has an analyzer that you can run first — it checks the selected disk and reports where the used space is and where the free space is. This tells you whether you need to defragment. Figure 5-42 shows the Disk Defragmenter utility in Windows XP after analyzing drive C. After the hard drive has been analyzed, you can also view a report that shows you detailed information about the analysis just performed. When you're happy with the analysis, you can perform the defrag by clicking the Defragment button. Figure 5-43 shows the report generated by the analysis of drive C.

Formatting partitions and volumes

After you have created the partitions or volumes, your next step is to format these partitions or volumes so that you may start storing data on them. When you format the partitions or volumes, which now show as drive letters in the My Computer icon, you choose which file system to format them with. Before you format the partitions, you should review the different types of file systems and the advantages and disadvantages of each (see the section "Understanding File Systems," earlier in this chapter). Formatting a drive prepares the drive for storing information. The format command creates a root directory on the disk as well as two tables used to store information about the files and to aid in the retrieval of these files. The first table is the Directory Entry Table (DET); it lists all the files stored on the drive, along with the date that each file was last modified. It also stores the starting cluster for each file stored on the drive. The DET is used when the system goes to open a file; the system looks in the DET for the file. Once an entry is found for the file being opened the starting cluster for the file is determined and then the system goes to that cluster to retrieve the file contents. The second table is the File Allocation Table (FAT). The FAT lists each cluster, showing you which clusters are used and which clusters are free. The FAT also indicates any clusters that have been marked as bad clusters, which are unusable. When a file spans multiple clusters, the FAT shows that the first cluster is linked to the next cluster by indicating the next cluster value as part of the FAT entry. The last cluster that is used to store the data for the file is marked with an end of file (EOF) marker — this is how the system knows it has reached the last cluster for the file. To format a partition/volume in Windows 2000/XP/Server 2003 using Disk Management, perform the following steps: 1. Click Start, right-click My Computer, and choose Manage. 2. When the Computer Management console has started, select Disk Management in the left side of the screen. 3. Over on the right side, in the bottom half of the screen, right-click the partition you want to format and choose Format,

Extended partition

An extended partition allows you to extend beyond the four-partition barrier by being a partition that contains one or more logical drives. A logical drive is a block of disk space that is assigned a drive letter. As an example on how you could use extended partitions, you could set up three primary partitions and then decide that you would like to divide the last chunk of free space into three additional parts (for a total of six partitions). If you create another primary partition out of some of the free space, then you will have four parts — and that is your limit, four partitions per disk. What you can do instead is create an extended partition out of the remaining space after the three primary partitions have been created and then create three logical drives inside the extended partition. Logical drives are not partitions, so you are not limited to four. This will give you your six desired parts. Book II Chapter 5 Working with Storage Managing Partitions and Volumes 243 An extended partition is, in effect, the space that remains after the primary partitions are defined. The extended partition does not have an actual drive letter assigned to it; it's simply a container that holds all the logical drives that you build. A logical drive is a logical division of the hard disk that the computer treats as if it were a separate disk drive; it's the actual area of the extended partition to which documents are saved. As an example, suppose you're partitioning a 6GB hard drive using the FAT file system. FAT cannot define partitions larger than 2GB, so you have to divide this drive into at least three different partitions: The first partition you define is the primary partition — a 2GB partition that also becomes the active partition (drive C). What's left is a 4GB extended partition that can store two logical drives (D and E), each of which can be no larger than 2GB.

Disk Cleanup

Another important tool for Windows XP is the Disk Cleanup tool. The Disk Cleanup tool scans a disk for files that can be safely removed from your system in order to free up disk space. The Disk Cleanup utility can remove a Figure 5-44: The Check Now option in Windows 2000/XP/ 2003. Book II Chapter 5 Working with Storage Understanding Management Tools 263 number of different types of files to help free up disk space. Disk Cleanup can remove temporary Internet files, Windows temporary files, and applications no longer used. To perform a disk cleanup, follow these steps: 1. Choose Start➪All Programs➪Accessories➪System Tools➪Disk Cleanup. 2. Select the drive you wish to clean up, as shown in Figure 5-45.

lba continued

As an example of why you take the lowest value in each category, if the hard disk only supports 4 heads then only 4 heads are detected. Although the BIOS supports a potential 16 heads, that doesn't mean they are actually there. Book II Chapter 5 Working with Storage Discovering IDE Devices 213 So the problem is that you have purchased a 2.1GB drive but the system is only recognizing 132MB! The solution to the problem is LBA or ECHS — again, both of these technologies offer the same solution. They were just built by different manufacturers. An LBAenabled BIOS can recognize 1024 cylinders, 256 heads, and 63 sectors — essentially being able to support more heads on the drive. As a result, the drive lies to the BIOS by using a translation factor of usually 2, 4, 8, or 16. The physical dimensions of the drive are taken and manipulated by the translation factor to calculate the logical dimensions that are reported to the BIOS. In our example, 16,384 cylinders are too many cylinders so they are divided by translation factor of 16 to reach the LBA maximum number of cylinders supported. To make up for the loss in cylinders, the heads are then multiplied by 16, ensuring that the logical number of heads falls under the LBA limit of 254. Table 5-2 shows the solution — notice that the size of the drive (2.1GB) is what the LBAenabled BIOS will recognize.

Termination

Both ends of the SCSI bus must be terminated so that when a signal is sent down the SCSI bus, it is absorbed at the end of the bus by the terminator. If the signal was not absorbed, or removed from the bus with a terminator, the signal would bounce back and collide with other data on the bus. A collision would destroy the signal. The first device in the chain must be terminated along with the last device in the chain, the first device usually being the host adapter. If the device is an internal device, terminating may involve modifying jumper settings. If the device is external, a terminator will be added to the back of the device. Figure 5-17 shows a terminator for external devices. If your SCSI chain has a combination of internal and external SCSI devices, then the card shouldn't be terminated because it is no longer the end of the SCSI chain. Instead, you should terminate the devices at either end of the SCSI chain. Most SCSI cards today are self-terminating so you typically will not need to terminate them manually yourself.

SCSI cabling

Different types of cabling are used to chain SCSI devices to the SCSI adapter. Internal devices use a 50-wire ribbon cable; external devices require a thick Centronics cable to connect to the Centronics 50 (typically used by SCSI-1) or the Centronics 68 (typically used by wide SCSI-2 technologies or Ultra SCSI-3) connector on the back of the device. The different versions of SCSI use a large number of different cable types. Figure 5-15 shows some internal SCSI connectors, and Figure 5-16 displays a handful of external SCSI connectors. Be sure to be familiar with these cable types for the exam.

DVD/DVDRW

Digital Versatile Disks (DVDs) are similar to CDs in the sense that they are another type of optical storage — but they store a lot more data. The typical DVD stores 4.7GB of data. Some DVDs (unlike CDs) can store data on both sides of the disk, and newer DVDs even store data on different layers on the DVD. This allows the DVD to store more than the 4.7GB, depending on the DVD standard. Table 5-5 describes the different DVD standards.

Addressing

Each device is assigned an internal address, a SCSI ID, in the SCSI bus. The SCSI controller knows the address of each device. When the SCSI controller receives information for a particular device, the controller references that device by its ID in the SCSI bus. This way, there's no confusion as to whom the data is destined for. You are responsible for assigning the SCSI IDs when you connect each device to the SCSI chain. You assign an ID either by jumpers or DIP (dual inline package) switches if the device is an internal device, or by a spinner if the device is external. A spinner is an indicator on the back of the external SCSI device whose value you can change by pressing the button to increase or decrease the SCSI ID. Figure 5-14 shows the back of an external SCSI tape drive and how to change the SCSI ID using the spinner.In Figure 5-14, you can also see the type of connector used for external SCSI devices. This is a Centronics 50-pin connector. If you are installing an internal SCSI device, you will most likely need to assign the SCSI ID by using jumpers. Internal SCSI devices have a jumper set with three pairs of jumper pins. The decimal values of these jumpers, although probably not shown on the drive, are 4, 2, and 1 (from left to right). Table 5-4 shows this jumper setup.

USB external drives

External drives are just as popular as flash drives today. Like flash drives, an external drive uses a USB connection. Unlike flash drives, external drives allow you to store hundreds of gigabytes, even a terabyte, of data. These are great solutions to add more space to a laptop computer.

Flash drives

Flash drives, also known as thumb drives or memory sticks, are the popular method for carrying data from computer to computer. Flash drives can store anywhere from 64MB to multiple gigabytes of information and are relatively cheap. Flash drives are USB devices that you simply plug into the USB port on the computer. Plug and Play kicks in, detects the device, and assigns the drive a letter in the My Computer icon. To access the flash drive, you simply doubleclick the drive in My Computer and open, copy, and move files as you wish.

LBA and ECHS

For the A+ exam you need to be familiar with Logical Block Addressing (LBA) and Extended Cylinder/Head/Sector (ECHS) and what their purpose is. Essentially LBA and ECHS perform the same goal — they perform sector translation; sector translation is the hard drive controller lying to the BIOS about the drive geometry. LBA was developed by Western Digital while ECHS was Segate's solution to recognizing larger drives. Keep reading to learn why you need sector translation. The reason you need sector translation is that the original BIOS code found on computers was limited to being able to see only 1024 cylinders, 16 heads, and 63 sectors — which is a total drive size of 504MB (1024 × 16 × 63 × 512). The problem is if you bought a 2.1GB hard disk, your BIOS would not recognize it because the geometry of the 2.1GB drive is too high for the BIOS. In this example the geometry of the drive is 16,384 clusters, 4 heads, and 63 sectors. By looking at Table 5-1 you can see that the lowest value in each category is what will be recognized by the system.

Managing Partitions and Volumes

For the A+ exam, you are required to know the steps to install a hard disk. After you have physically connected the drive and configured the jumper settings, you need to be aware of the steps to configure the partitions on the disk. The following is the order in which you configure the partitions on the disk: 1. Create a primary partition. 2. Create an extended partition. 3. Create a logical drive in the extended partition. 4. Format the drives to create a file system. Understanding the order of the steps for partitioning and formatting a disk is important for the A+ Certification exam. Be sure to memorize the above list in order to correctly answer exam questions about preparing the hard disk. 242 Managing Partitions and Volumes A partition is defined as a segment of the hard disk, created by dividing the disk logically into discrete units. You create partitions for a number of reasons — you may partition a disk to organize your applications and operating system on drive C while storing your data on drive D. You may also partition a disk for more technical reasons, such as to run multiple operating systems on the same machine. Whatever the reason for creating a partition, how you create and manage partitions is important for the A+ exam. This section examines different types of partitions and provides steps for creating, deleting, and formatting them. Frequently, a partition is a means of providing better access to the information stored on a disk. For example, telling the kids that their games are on the D drive is usually easier than describing a complex path to the folder that holds the games. You are limited to four partitions per disk, so be sure to plan them carefully. Operating systems such as DOS, Windows 9x, and Windows 2000/XP/2003 (that are using basic disks — more on that in the section "Creating partitions and volumes in Windows 2000/XP/Server 2003," later in this chapter) can create two types of partitions: primary partitions and extended partitions.

installing ide

For the exam, remember that an IDE hard drive uses a 40-wire ribbon cable, while a floppy drive uses a 34-wire ribbon cable. Also note that Ultra DMA uses an 80-wire ribbon cable that contains 40 wires for data and 40 additional grounding wires. One of the wires (known as Wire 1) on the IDE ribbon cable is a different color than the others; usually, it's red, but it may be blue. Wire 1 must be placed over Pin 1 when connecting the ribbon cable to the hard drive and motherboard — a procedure known as the Pin 1 rule. The big question is: How do you know which pin is Pin 1? Hopefully, the manufacturer has indicated Pin 1 by placing a small "1" near it. If you look at the connector on the hard drive and on the motherboard, you may see a small "1" on one end of the connector. That's where you need to place Wire 1 when connecting the ribbon cable. Sometimes the manufacturer will place a "40" by Pin 40 instead of displaying where Pin 1 is, so watch for that as well. Sometimes the manufacturer may not have enough space to indicate Pin 1 and so does the opposite, which is to indicate Pin 40. This method gives you the same information, though: If you know what side Pin 40 is on, you know that Pin 1 is on the opposite side. After you have connected the IDE ribbon cable, you want to give the hard disk power from the power supply so that you can run the motor in the drive. (People often forget this step and then wonder why the drive doesn't work.) Figure 5-8 shows how to connect the ribbon cable and the power supply cable to the hard disk.

The Serial ATA Hype!

IDE technology has been around for many, many years, and there has been a big need for a change in hard drive technology — that change came as a new hard drive interface called Serial ATA (SATA). IDE is a parallel technology, and though SATA is a serial technology, it offers great speed and other benefits. SATA is also a lot faster than IDE — approximately 30 times faster, with current speeds of 150 Mbps and future speeds of 600 Mbps. One of the first benefits of SATA is that it is a hot-swappable technology, meaning that you can add or remove drives from the system without shutting the system down. This is a huge benefit when you look for RAID solutions for servers that you don't want to spend a lot of money on — like a server for a small company. (For more on RAID, see the section "Securing Data with RAID," later in this chapter.) Another benefit of SATA is in the cabling. Because SATA is a serial technology, the cables can be longer than your typical IDE ribbon cables. I don't known how many times with IDE I had to switch the CD-ROM and my second hard drive around just so the ribbon cable could reach. SATA cables can be 39 inches long, while the maximum distance for IDE is 18 inches. The other benefit of the cabling with SATA is that it uses only 7 wires, as opposed to the 80 wires used in newer IDE drives. The benefit here is that it allows for better airflow in the system, which results in a cooler system If you don't have a motherboard that has SATA connectors on it, then you can get a PCI card that does have the connectors. Also, for backward compatibility, you can get a SATA bridge that allows IDE drives to be connected to a SATA system.

Installing SCSI devices

If you understand the issues with SCSI, installing a SCSI bus is fairly simple. First, you want to assign a unique ID number to each device. I usually perform this step at the beginning so that when everything is connected, I won't have to play around figuring out how to change the ID of the devices. For more on assigning the IDs, see the section, "Addressing," earlier in this chapter. When assigning the ID numbers, remember that you want to assign the bootable drive the ID of 0 because the SCSI host adapter automatically looks to SCSI ID 0 for a bootable device. You also want to enable the SCSI BIOS on the SCSI controller if you are booting off a SCSI hard disk. When you enable the SCSI BIOS, you won't need to install a driver for the card because the PC will recognize the device on startup. If you are booting off an IDE drive and using the SCSI disk as an additional drive, you should disable the SCSI BIOS and install a driver in the operating system. After you assign the ID numbers to each device, insert the SCSI host adapter into the expansion slot of the PC. After inserting the SCSI card, chain all of the devices together. When preparing for the A+ exam, it is important to know the different terms used for a particular technology. Another term for a SCSI chain is a daisy chain. When you chain the devices together, you are creating a daisy chain. When you have the devices chained together, make sure that each end on the SCSI bus is terminated. If the last device is an external device, you need to put the terminator on the end of the device. If you are installing internal devices, you need to check the documentation on the internal devices to find out what jumpers to set. At this point, the IDs are configured for each device, the SCSI card is inserted into the PC, and the devices are connected to the card to create a SCSI bus. Book II Chapter 5 Working with Storage Learning How SCSI Works 231 We also terminated the SCSI bus at either end. Before installing the driver for the SCSI card in the operating system, I'll review the steps to install a SCSI device one last time. To install a SCSI bus: 1. Assign unique IDs to each device. 2. Install the SCSI host adapter into the expansion slots. 3. Chain devices to the SCSI host adapter. 4. Terminate the SCSI bus at both ends of the chain. 5. Install the driver for the SCSI card if you are not booting off the first hard disk. After you have connected the hardware for the SCSI chain, you need to load a driver in the operating system for the SCSI host adapter. After inserting the SCSI card and powering on the Windows operating system, Plug and Play should kick in, asking for the driver of the new hardware. If Plug and Play does not kick in, you can run the Add Hardware Wizard found in the Control Panel or My Computer properties of the system. To install a SCSI adapter in Windows 2000/XP, and Windows Server 2003, follow these steps: 1. Choose Start➪Settings➪Control Panel in Windows 2000 or right-click My Computer and choose Properties in Windows XP and Windows Server 2003. 2. Double-click the Add Hardware icon in Windows 2000. If you are using Windows XP, go to the Hardware tab and click the Add Hardware Wizard button.

master/slave

Master/slave configuration If you are installing multiple IDE devices, you are creating an IDE chain. The chain will be made up of one 40-wire IDE ribbon cable with two drives connected to it. Most IDE ribbon cables today have three IDE connectors on them — one that connects to the motherboard and one for each of the two drives that can be connected in a single chain. After you have the two drives connected together, you need to configure the drives into a master/slave configuration. Why this type of configuration? Because each drive has a built-in controller that makes it act like its own boss. Have you ever tried to work in an environment with two bosses? The purpose of designating a master is to specify who the boss of that chain is — the controller that will be responsible for communicating with the processor. When setting up a multi-drive system, you have two drives, each with a controller that can potentially send and receive signals to and from the processor. To save confusion, one of the drives is designated as the master. The master drive receives all signals from the processor and sends back any data on behalf of both drives. The other drive is designated as the slave. The slave drive passes any information it wants to send to the processor up to the master, which then forwards that information to the processor.

NTFS

NTFS Starting with Windows NT, Microsoft implemented a new file system called New Technology File System (NTFS). NTFS makes better use of the space available on a particular disk by using 512 bytes as the cluster size, which is the same size as a sector! This means that you are wasting even less space on an NTFS file system than on a FAT32 file system. The original version of NTFS supported a number of features that made it more attractive than the FAT versions of the file systems. With NTFS, you could configure permissions that controlled who could access what files. You could also take advantage of features such as compression and auditing. One of the biggest complaints with the original version of NTFS is that it had no way to limit how much disk space a user could use. As a result, users could waste gigabytes of hard disk space on the server, and the administrator could not stop the user unless a third-party program was purchased. Limiting disk space usage is one of the improvements that Microsoft made on the next version of NTFS, known as NTFS version 5.0, which was implemented with Windows 2000 and every Windows OS after that.

more

Note that the extended partition itself has no drive letter assigned to it. The extended partition is just a container to hold the logical drives — and they take the drive letters. Users of the system will be able to store data on drive C, drive D, or drive E! A hard disk can contain no more than four partitions, only one of which can be the extended partition. This means you could have three primary partitions and one extended partition to hold any logical drives. Having three primary partitions also shows why you have to set the active partition. A primary 2GB (Drive C:) 4GB Extended Partition 2GB (Drive E:) 2GB (Drive D:) Primary Partition (Active) Logical Drive Logical Drive Figure 5-21: Partitioning a hard disk. 244 Managing Partitions and Volumes partition is a bootable partition. But if I have three primary partitions, which one do I boot from? The answer is simple — the one defined as the active partition. Note that when you create the partition during the installation of Windows the partition is automatically marked as being the active partition.

ide exam question

On the A+ exam, you may be asked the number of devices that IDE supports. IDE supports two devices, while EIDE supports four devices. Although there have been improvements in the IDE technology, Table 5-3 displays some of the original specifications.IDE has been around for quite some time now, and as a result has gone through some changes. The following sections outline some of the technologies and terms that describe the different versions of IDE. Be sure to know these for the exam. IDE/ATA standard A number of hard drive standards have been developed over the last two decades — the first major standard being the IDE standard. The Integrated Drive Electronics (IDE) standard, which has been around since 1989, calls for an integrated controller on the drive to manage information entering and leaving the hard disk. IDE drives attach to the motherboard by means of a 40-wire ribbon cable. The IDE standard also allows two drives to daisy-chain, creating a master/ slave relationship between devices. The master drive is responsible for sending and receiving information in the chain. The IDE standard is also known as the Advanced Technology Attachment (ATA) standard, which is sometimes known as the ATA-1 standard.

ATAPI specification

Originally, IDE devices were implemented as hard drives, but an additional ATA specification allows other types of devices to exist on an ATA (or IDE) chain. This specification is known as the ATA Packet Interface (ATAPI), which 216 Discovering IDE Devices allows devices like CD-ROMs and tape drives to exist on an ATA chain. Other types of ATAPI devices are CD writers, DVD devices, and zip drives.

Types of SCSI

Over the last twelve years or so, SCSI technology has increased in performance to stay competitive with advances in IDE and EIDE. Newer versions of SCSI have amazing transfer rates, which is one of the reasons why you find network servers using SCSI hard drives instead of EIDE devices. The following list outlines the key points about the different versions of SCSI: ✦ SCSI-1: The original version of SCSI, SCSI-1, was an 8-bit technology with a transfer rate of 5 MBps. One of the major benefits of SCSI was that you weren't limited to two devices in a chain like you are with IDE. SCSI-1 allowed you to have eight devices in the chain, with the controller counting as one. ✦ Fast SCSI-2: Fast SCSI-2 increased the performance of SCSI by doubling the transfer rate. Fast SCSI-2 devices transfer information at 10 MBps. Fast SCSI-2 is still an 8-bit technology and supports eight devices in the chain. ✦ Wide SCSI-2: Wide SCSI-2 doubled the 8-bit data path of SCSI to 16 bits. Doubling the width of the data path raised the transfer rate to 10 MBps, like Fast SCSI-2, but Wide SCSI-2 can support 16 devices in a chain. When trying to remember the difference between SCSI-1, Fast SCSI-2, and Wide SCSI-2, think of it this way: fast implies speed, so the transfer rate is increased. Wide implies "wider" or bigger, which is the data path that has been increased; as a result, you also get a higher transfer rate. ✦ Fast Wide SCSI-2: Fast Wide SCSI-2 is the combination of Fast SCSI-2 and Wide SCSI-2. The data path of Fast Wide SCSI-2 is 16 bits, the transfer rate is 20 MBps, and it supports 16 devices in a chain. Ultra SCSI: Ultra SCSI takes the transfer rate of 10 MBps and doubles it again to 20 MBps! With Ultra SCSI, the bus width is only eight bits, and the number of devices that exist in the chain is eight. ✦ Ultra Wide SCSI: Ultra Wide SCSI is Ultra SCSI with the bus width increased to 16 bits, and the number of devices in the chain is increased to 16. The transfer rate of Ultra Wide SCSI increased to 40 MBps. ✦ LVD (Ultra2 SCSI): Low Voltage Differential (LVD), also known as Ultra2 SCSI, has a bus width of 16 bits and supports up to 16 devices. LVD is a popular SCSI version due to having a high transfer rate of 80 MBps. Be prepared to answer questions on the A+ exam about the different types of SCSI. I suggest memorizing the transfer rates of each type.

Read/write process

Platters are divided into 512-byte sectors. These sectors are the area on the platter that data is written to. The platters have a magnetic coating applied to them that is extremely sensitive to magnetism. While the platters spin, the read/write head moves from track to track until it reaches the desired track. Then it waits for the appropriate sector to move underneath it, at which time the read/write head is energized to apply a magnetic charge to the particles in the disk coating. This changes the particle binary state from zero to one, thus creating data. The same happens when the data needs to be read, the read/write head moves over the appropriate sector and reads the data that resides in the sector. Understanding Hard Drive Terminology 211 The read/write heads don't actually touch the surface of the disk platters; instead, they hover about ten micro-inches (or millionths of an inch) above it — that's not even enough space to place a hair between the read/write head and the platter's surface. This design helps improve disk performance because a read/write head that makes contact with the platter would cause friction, slowing down the rotation speed of the disk and creating extra heat.

SCSI overview

SCSI is an acronym for Small Computer System Interface. The important part of this term is small computer, meaning that SCSI has its own brain, known as the SCSI adapter, that handles the SCSI environment. This SCSI adapter (also Master J20 Slave No Jumper J20 Jumper Figure 5-11: Looking at an additional jumper configuration for a dualdrive setup. Book II Chapter 5 Working with Storage Learning How SCSI Works 223 known as a SCSI card or SCSI controller) is responsible for managing all SCSI devices and controlling the conversation on the SCSI chain. SCSI technology has many advantages over IDE technology, such as ✦ The types of devices supported: SCSI supports a multitude of devices, including hard drives, CD-ROMs, scanners, printers, and tape drives, to name just a few. This is a huge benefit because originally IDE typically only supported hard drives and CD-ROMs. ✦ The number of devices supported in a single SCSI chain (also known as a SCSI bus): Original versions of SCSI supported up to eight devices in the chain, but one of those devices is the SCSI card that's added to the computer to give you the capability to use SCSI. Remember that IDE only allows two devices in the chain, and EIDE supports four devices, so with SCSI you are not only allowed to have more types of devices, but you are also allowed to have more of those devices! Original SCSI supports up to eight devices in the chain, but if the exam asks how many devices can be attached to a SCSI adapter, the answer is seven. If the test asks how many devices can exist in the SCSI chain, the answer is eight — the card counts as one of those devices. Remember to watch the wording of the questions closely. ✦ The performance of SCSI over IDE devices: Original SCSI devices don't compare in the performance category with EIDE devices, but some of the later SCSI technologies, such as SCSI-2 and SCSI-3, can outperform IDE and EIDE. I give details about transfer rates of the different types of SCSI devices in the "Types of SCSI" section, later in this chapter. For the remainder of this chapter, assume that you are dealing with original SCSI, also known as SCSI-1. In the "Types of SCSI" section, I discuss the newer types of SCSI.

Sectors

Sectors The platter is divided into pie-shaped slices, called sectors. Now the confusing thing about sectors is that where a track intersects with a sector creates sector blocks — also known as sectors! Each sector (block) is 512 bytes in size. and is the actual storage area for data. Figure 5-3 illustrates the tracks and sectors on a disk platter. Each pie-sliced sector has an address; the first sector is known as sector 1, the second sector is known as sector 2, and so on. The reason why the term sector applies to two different areas of the disk is because if you were to look at a sector (the pie slice), like the pie slice labeled in Figure 5-3 as "sector," it may be sector one. Now all of the sector blocks in that pie slice are all considered sector one as well, but they differ in the track that they reside on. So each sector block has an address that is made up of the platter side number, the sector, and track number. For example, data can be saved to side 1, sector 2, track 4 — which is the address of a 512-byte sector block. Note that the term sector block is a term I made up for this discussion; the term sector is also used to describe the 512-byte blocks.

The FAT file system

The File Allocation Table (FAT) file system has been the most popular file system up until the last few years. Although the FAT file system is the most common (it can be used by all operating systems), it is losing the popularity contest to its successor — FAT32 — due to its age and limitations. The FAT file system was the file system used by DOS, Windows 3.1, and Windows 95, and is supported in Windows NT, Windows 98/Me, and Windows 2000/XP/2003. FAT's biggest strength is that it's the file system most widely understood by different operating systems — but it has many shortcomings. One of the major shortcomings is that it cannot create a partition larger than 2GB. (A discussion of partitions is coming up in the "Managing Partitions and Volumes" section; for now, consider a partition simply as a discrete portion of space on the disk.) The 2GB size limit was not a major limitation until hard drive sizes exceeded a few gigabytes. For example, a problem with the FAT file system is that a 20GB drive would need to be divided into 10 partitions to use all the space — an impractical and inappropriate use of space. Can you imagine being required to divide your home up into ten different 238 Understanding File Systems rooms whether you wanted to or not? Instead of five spacious rooms, you'd get ten cramped rooms! Not practical! Earlier in this chapter, in the section "Disk geometry," you discover the characteristics of a disk, including clusters, which are groups of sectors (each sector taking up 512 bytes on the disk). To refresh your memory, a cluster is what a file is written to, and only one file can occupy a cluster at a time. The cluster size is determined by the partition size and the file system being used. For example, you may have a 2GB FAT partition with a 32K cluster size. The issue with clusters is that if you have a 32K cluster size and you save a 12K file, then you waste 20K of hard disk space because only one file can occupy a cluster. Over time, as more files are saved, this could add up to a lot of wasted space! The solution is to use smaller partition sizes, which create smaller cluster sizes, or to use a different file system that uses smaller cluster sizes. Bottom line, the FAT file system uses clusters inefficiently. Table 5-6 lists the cluster sizes used with different partition sizes on FAT file systems.

Master Boot Record

The Master Boot Record (MBR) is the first sector on the first track of the first side of the first platter; it holds the operating system boot code that controls the loading of the operating system. The MBR also holds drive characteristics, such as the partition table. During the boot process, the system has to find a primary partition that is active — it does this by looking at the partition table in the master boot record. In general, if anything goes wrong with the MBR, you won't be able to boot the system. Because the boot record is always in the same location on every disk, it becomes very easy for a malicious hacker to write viruses that modify or corrupt the MBR. This is one reason you should always run virusdetection software. Refer to Book IX, Chapter 3, for more on virus detection and protection.

ide overview

The hard drive controller is responsible for converting signals made by the system CPU to signals that the hard disk can understand. These signals 214 Discovering IDE Devices include instructions on where to find data and how to get to that data. The hard disk performs its task, and any data that needs to be returned is sent to the controller from the hard disk. The controller then converts the signals from the hard disk into signals that the system can understand. In the past, the controller was on an expansion card, which was connected to the drives via ribbon cables. The goal of IDE was to make the installation of hard disks easier by including the controller on the hard disk, which is where the name comes from: Integrated Drive Electronics (IDE). So, today's drives have the controllers integrated into the drives themselves, meaning the drive is its own boss. Originally, IDE was available only in the flavor of hard drives; originally you did not have any other type of IDE device such as CD-ROM or tape drives. IDE hard drives had a maximum capacity of about 528MB. Another important limitation with IDE is that only two devices could be connected in a chain. Back in the day when IDE was first used, SCSI was better in that respect: It supported eight devices in a chain (more on SCSI in the "Learning How SCSI Works" section of this chapter). Original IDE devices have a transfer rate of about 10 Mbps and may have cache on the drive itself. The cache memory is a small amount of memory for storing data that is used frequently to increase drive performance. Due to its limitations, IDE has been replaced by Enhanced Integrated Drive Electronics (EIDE). EIDE devices have a transfer rate of about 16 MBps. Four devices are allowed in an EIDE chain, with a greater variety available. For example, you may now add CD-ROMs and even zip drives to the EIDE chain. Note that the capacity of the drives has been dramatically increased with EIDE — it now supports drives of over 200GB! Table 5-3 compares the features of IDE and EIDE.

RAID 5 Volume (RAID level 5)

The problem with RAID level 1 is that you essentially waste half of the money you spend on hard drives because, under normal conditions, you only use one disk. A RAID solution that you can use where you get more disk space for your dollar is known as RAID level 5, or a RAID 5 volume. Microsoft's implementation of RAID 5 volume uses a minimum of 3 disks and can use up to 32 disks. When data is saved to the RAID 5 volume, the information is written across all disks in the array. A RAID 5 volume also stores parity information on a different disk for each write operation. This parity information is an "answer" that is generated after the data being written is run through an algorithm. The answer is then stored on one of the disks for that write operation. If a disk fails and a piece of data cannot be retrieved, the data is recalculated based on the answer (parity data) that is stored in the array. Sound confusing? Here's a simple example: What is the value of x in the following formula? 4 + x = 9 You probably had no problem coming up the answer of x = 5. You simply subtracted the known data from the given answer to calculate what is missing. RAID 5 volumes do the same thing with the parity data (the "answer") when there is a disk failure. When one disk fails, the fault-tolerant disk driver simply subtracts the known data (from the existing disks) from the parity data to generate the missing data on the fly. A popular question I get is this: "Will RAID 5 volumes be able to recover from multiple disk failures?" To answer that, look at a simple algebra equation. What is the value of x in the following formula? y + x = 9 We cannot say for sure what the value of x is because there are now two variables involved. The software implementation of RAID that is built into the operating system can calculate the missing data only when there is one failed drive. Now take a look at how to create a RAID 5 volume in Windows Server 2003. Remember that a RAID 5 volume must use at least three disks — so be sure to have three disks with unallocated space. Then follow these directions: 1. In the Disk Management console, right-click unallocated space and choose New Volume. 258 Securing Data with RAID 2. When the welcome screen to the New Volume Wizard appears, click Next. You are asked what type of volume you wish to create. 3. Choose RAID 5 (as shown in Figure 5-39) and then click Next.

Mirroring/duplexing (RAID level 1)

The type of hard drives that are normally found in servers are SCSI drives, which means that there is a SCSI adapter (controller) that connects the drives to the systems. Disk mirroring is the use of two disks on a single controller to create full redundancy — whatever is placed on one disk is copied to the second disk. When creating a disk mirror, you must use two disks so that if one disk fails you can rely on the copy of the data that is stored on the other disk. When the mirror is established, you will have a new drive letter that's accessible from the My Computer icon. This drive actually shows the data stored on both disks; if you save a file to this drive, it's written to both disks that make up the mirror; you don't see two representations within the My Computer icon. But if you use the Disk Management console, you see that the two disks are part of the mirrored volume. Disk duplexing is the same idea as disk mirroring but requires the installation of an additional SCSI controller. Disk mirroring is fault-tolerant if a drive fails (because the other drive is available), but there is no fault tolerance if the controller fails (because there is only one). If you add an additional controller and place one drive on one controller and the other drive on the other controller, you have a more fault-tolerant solution. If one drive fails, you have the other; if a controller fails, you have the other drive running off the other controller. Figure 5-33 shows the difference between a mirror and a duplex. To create a mirrored volume on Windows Server 2003, you first need to ensure that you have converted both of the disks that you want to use as mirrors to dynamic disks. See the previous section, "Dynamic disks," to find out how to do this. After you have converted both of the disks to dynamic disks, you can create a mirrored volume by following these steps: 1. Click Start, right-click My Computer, and choose Manage. 2. In the Disk Management console, right-click the unallocated space and choose New Volume 3. When the welcome screen to the New Volume Wizard appears, click Next. You are asked what type of volume you wish to create. 4. Choose Mirrored and then click Next (as shown in Figure 5-35).

NTFS 5.0

This newer version of NTFS has a few extra features over original implementations of NTFS, one of which has been long overdue — disk quotas. Disk quotas allow the system's administrator to choose the amount of disk space that each user is allowed to use by placing a limit on the disk. For example, when managing the home directories, you can ensure that Bob is not allowed to use more than 500MB of disk space. Another feature of NTFS 5.0 is the Encrypting File System (EFS). EFS uses public key/private key technology to encrypt a file stored on the hard drive. When a file or folder is encrypted with EFS, only the person who created the file or the recovery agent (by default, the administrator is the recovery agent) can open the file. When using EFS, even if another user has permission to view the file, he or she will be unable to do so because the file is Book II Chapter 5 Working with Storage Managing Partitions and Volumes 241 encrypted. The encrypting file system is a big selling point for organizations with mobile users who need to protect the privacy of the data that sits on their laptops. To summarize, the NTFS file system offers the following features over FAT and FAT32: ✦ The capability to secure the resource through permissions ✦ The capability to secure files through encryption ✦ The capability to enable auditing to monitor who accesses the files and folders ✦ The ability to compress the contents of a file or folder

Creating partitions and volumes in Windows 2000/XP/2003

To create partitions in Windows 2000/XP/Server 2003, you use a partitioning tool known as the Disk Management snap-in. You can open the Disk Management snap-in by choosing Start➪Control Panel, clicking Performance and Maintenance, clicking Administrative Tools, and double-clicking Computer Management. Then, in the Computer Management window, select Disk Management. Figure 5-22 shows the Disk Management snap-in from Windows 2000/XP/Server 2003. When managing a disk in Windows 2000/XP/Server 2003, be aware that there are two types of disks: basic and dynamic. Basic disks Basic disk is the term Microsoft uses to describe a disk that supports partitions and all the limitations of partitions. If you can create partitions on a disk in Disk Management, then you are working with a basic disk. A basic disk has the following characteristics: The disk is divided into partitions ✦ You are limited to four partitions per disk ✦ You are limited to one extended partition per disk ✦ You create primary partitions, extended partitions, and logical drives ✦ A basic disk is the default disk type in Windows 2000/XP/2003 IT professionals have become accustomed to these characteristics when preparing disks for DOS, Windows 9x, and Windows NT. With Windows 2000 and above, you still have these limitations when working with a basic disk, but you may convert the basic disk into a dynamic disk. The benefit of converting to a dynamic disk is that you no longer work with partitions, so there are no partition limitations! You find out more about dynamic disks in the next section, but first you need to see how to create partitions on basic disks. The following steps demonstrate how to create a partition on a basic disk within Windows XP. The steps are similar in Windows 2000 and Windows Server 2003. 1. Click Start, right-click My Computer, and choose Manage. 2. When the Computer Management console has started, select Disk Management on the left side of the screen. On the right side, on the bottom half of the screen, notice that the disk type for Disk 1 is Basic Disk. 3. Right-click the unallocated space and choose New Partition, as shown in Figure 5-23. The New Partition Wizard starts and displays a welcome screen. 4. Click Next. The wizard asks what type of partition you will be creating, as shown in Figure 5-24. 5. Select Primary Partition and click Next. 6. Type the desired size of the primary partition in megabytes and click Next. 7. Choose the drive letter you want to associate with this partition and then click Next. 8. Choose the file system you want to format the partition as (shown in Figure 5-25) and then click Next. 9. Click the Finish button. You will notice that the drive starts to format in the background, and it indicates the percent complete in the Disk Management utility. 10. When Disk Management is finished formatting the drive, close the Computer Management utility.

Check Disk utility

When a drive is formatted, two FAT tables are created — the Directory Entry Table and the FAT table. The FAT table links all the clusters that make up a file. What if that link is lost, or points to the wrong location? If the file system loses the link that joins two clusters together, the file becomes unreadable and has lost its data integrity. ScanDisk is a Windows utility in older Microsoft operating systems that has been replaced by the Check Disk option in Windows 2000, XP, and Server 2003. The Check Disk option not only looks for lost links but also scans the disk surface for bad blocks (clusters that cannot be written to), marking them as bad so they are not used. To perform an integrity check on a drive in Windows XP, follow these steps: 1. Open the My Computer icon and select the drive you want to check. 2. Right-click the drive and choose Properties. 3. On the Tools tab, click the Check Now button. Figure 5-44 shows the dialog box that appears. You may choose to automatically fix any file system errors and specify whether you want Windows XP to scan the surface of the disk and attempt to fix any problems with sectors. 4. Choose the options you want and then click Start. The scan may take a few minutes and when completed will display summary information.

MORE

When the Format dialog box appears, you may choose from a number of settings (shown in Figure 5-32) on how you wish to perform the format. The settings include • Volume Label: Use this option to specify a label for the partition. For example, I usually have a drive that is labeled "Data" where I store all my data. This label is just a friendly name used to identify what you might be using the drive for. • File System: Use this option to specify whether you want to use FAT, FAT32, or NTFS as the file system. Allocation Unit Size: Use this option to select the cluster size for the partition or drive. • Perform a Quick Format: A quick format does not perform a surface scan on the disk to check for errors. • Enable File and Folder Compression: You may implement compression on the drive by selecting this option. Compression allows you to save disk space by using less space on the disk for the data that is stored there. The compression feature shrinks the file when the file is saved and decompresses it when the file is opened. 4. 4. Select and/or enter your options and click OK to format the drive. 5. Click OK to format the drive. You may also use the command-line version of the format command by going to a command prompt and typing format. For example, to perform a quick format of drive D and then apply a volume label of "Pictures," the command is: format d: /q /v:Pictures Labs 5-3, 5-4, and 5-5 give you the opportunity to practice creating partitions and volumes in Windows XP. You can find these labs in the Labs.pdf file in the Author directory of the CD-ROM. Securing Data

Host adapter

When you install SCSI devices, you first need to install the SCSI host adapter. The SCSI host adapter is an expansion card that you add to the computer so you can chain SCSI devices off the adapter. In essence, the SCSI host adapter is the brain of the SCSI bus; it acts as the controller for the SCSI bus. Figure 5-12 shows the SCSI adapter being inserted into an expansion slot in the system. The SCSI controller (adapter) is responsible for sending and receiving all information to and from the SCSI bus, just like the IDE controller. When the system has information for one of the devices in the SCSI bus, the system hands the information over to the SCSI controller, which then passes the 224 Learning How SCSI Works information to the appropriate device in the chain. Figure 5-13 shows a SCSI bus — made up of the SCSI host adapter (the SCSI controller) — along with two internal SCSI devices and two external SCSI devices. The beauty of the entire setup is that the SCSI adapter in the computer is assigned resources, such as an IRQ and an I/O address. Each device in the SCSI chain is not assigned these resources because all processor information passes to the SCSI controller, and the controller passes the information to the devices. This means that the system never talks to the devices directly, so each device does not require an IRQ and an I/O address. When you go out to purchase a SCSI adapter, you first have to look inside your system to figure out what type of expansion slots are free. Today, you will typically have some PCI slots, but you may have an ISA or an EISA slot, as well. The difference between these expansion slots is performance — PCI runs at 33 MHz, while ISA and EISA run at only 8 MHz. Also, PCI and EISA are 32-bit technologies, while ISA is only 16-bit. The bottom line is that if you have some PCI slots free, you will probably end up purchasing a PCI SCSI adapter. So the next big question is when the SCSI controller receives information for a particular device in the chain, how does it send the information to that device?

Securing Data with RAID

Windows servers have a built-in software implementation of RAID (Redundant Array of Inexpensive Disks) that you can take advantage of. RAID is a method of implementing redundancy — duplicated information — on your hard drives; if one disk fails, the other disk(s) can provide the missing information. RAID is a method to implement fault tolerance. Fault tolerance is the idea that if there is a hardware failure in the system, such as if a hard drive fails, the Figure 5-32: Format options available in Windows XP when formatting a drive. Book II Chapter 5 Working with Storage Securing Data with RAID 253 system can continue to operate as normal. There are many different levels of RAID, but the only RAID levels that provide redundancy in Windows Servers are RAID level 1 and RAID level 5. Windows Servers also support RAID level 0, called disk striping or striped volumes. No redundant information is stored on striped volumes, which is why striped volumes are classified as RAID level 0. (Think of it as a zero level of redundancy.)

HPFS

Years ago, the High Performance File System (HPFS), which gained its popularity with the OS/2 operating system, was a major improvement over the FAT file system. Some of the benefits of OS/2 are that it supports long filenames, up to 254 characters (including the path). HPFS also supports partition sizes up to 2000GB and uses a cluster size of 512 bytes! When looking at the benefits of HPFS, you may be thinking, "What's the big deal, I get that with FAT32?" The big deal is that HPFS was released well before FAT32, or even before Windows 9x was designed. The disadvantage of HPFS is that it is not widely supported. Operating systems such as DOS and Windows cannot access HPFS volumes.