FIT-3 Projet 1 - Architecture des Ordinateurs

Annoncé: 06.11.2001

Due: Jeudi 06.12.2001

 

Voici des sujets de votre premier projet: Compréhension et présentation 

(il y aura 2 au total ce semestre).

 

Il s'agit de projets de présentation de l'état d'art (state of art) dans le domaines respectifs. La source principale de vos recherches sera la toile WEB, ainsi que les chapitres correspondants de livres de Tanenbaum et Stallings (disponibles après

les horaires de cours). La mise en page et à vous, et une proposition concernant la structure par chapitres est donnée en dessous (désolé … en anglais seulement).

 

Les versions définitives de vos travaux doivent représenter une ensemble compact par sujets, c.à.d. à la fin on doit obtenir 4 "grands" projets 1., 2., 3. et 4. que vous allez échanger entre vous.

 

3. Disques durs (Groupes 6, 7 et 8)

3.1. Aspects physiques, Organisation et Formatage, Spécifications

3.2. disques IDE, EIDE, SCSI

3.3. disques RAID

 
Hard Drives

http://www.pcmech.com/hdindex.htm

As anyone who has worked with a computer will know, a hard drive is one of the most important parts of your computer. Although it is a complex piece of hardware and installing one can sometimes be a chore, a little knowledge and practice can make it very easy. The process of replacing or installing hard drives can make you very nervous, but it doesn't have to.

 

This section of the site is written to help you decide on various types, learn how they work, and explain how to install a drive and conveniently prepare it for use.

 

1.            Hard Drives- Internal Parts

2.            File Structure

3.            Hard Drive, Heal Thyself

4.            Installing A Hard Drive - Step by Step

5.            HDD Configuration

6.            Partitioning

7.            Formatting

8.            ST-506/412 & ESDI Interface

9.            IDE Interface

10.        SCSI Interface

11.        SCA SCSI Drives

12.        IDE vs. SCSI

13.        Ultra ATA/66

14.        ATA/66 vs. ATA/33

15.        Ultra ATA/100

16.        RAID: Your Guide

17.        Common Hard Drive Problems

Hard Drives- Internal Parts

http://www.pcmech.com/show/harddrive/65

 

All hard drives share the same basic structure, varying only in how each part is used and the quality of the parts themselves. The platters, spindle motor, heads, and head actuator are inside the drive, sealed from the outside. This chamber is often called the head disk assembly (HDA). The HDA is rarely opened, except by professionals. On the outside are the logic board, bezel, and mounting equipment. Below, I will describe each of these components.

The platters are the disks inside the drive. Platters can vary in size. Often the size of the drive, 5.25" or 3.5", is based on the physical size of the platters. Most drives have two or more platters. The larger capacity drives have more platters. They are usually made of an aluminum alloy so that they are light. The newest and largest drives make use of a new technology of glass/ceramic platters. Basically, this is glass with enough ceramic within to resist cracking. This glass technology is taking over aluminum in the hard drive industry. Many popular manufacturers already use it, including Maxtor, Toshiba, and SeaGate. Glass platters can be made much thinner than aluminum ones, and they can better resist the heat produced in operation.

 

Alone, platters are not capable of recording data. Each one is coated with a film of some magnetically sensitive substance. The oxide media is one of the ways of doing this. A mixture of compound syrup is poured on the platter, then spun to evenly distribute the film over the entire platter. This substance has iron oxide as a main ingredient, explaining why many platters you may see will be brownish-orange. The other main media consists of a thin film of a cobalt alloy which is placed on the platter through electroplating, much like chrome.

 

The read/write heads do just that, they read and write to the platters. There is usually one head per platter side, and each head is attached to a single actuator shaft so that all the heads move in unison. Each head is spring loaded to force it into the platter it reads. When off, each head rests on the platter surface. When the drive is running, the spinning of the platters causes air pressure that lifts the heads ever-so-slightly off the platter surface. The distance between the head and platter is very small...so small that the HDA must be assembled in a clean room because one dust particle can throw the whole thing off. This sensitivity and accuracy is what causes only bigger companies to be able to repair hard drives simply because of the expense of a clean room. A slider is attached to each head. This mechanism actually glides over the platter and holds the head at the correct distance to do its job.

 

The head actuator is the device that all the heads are attached to. This part is in charge of moving the heads around the platters. They come in two types: stepper motor actuators and servo motor actuators. The stepper motor design is actually an electric motor that moves from one stop position to another, governed by click stop positions. They cannot stop between stop positions. The motor is small and is located outside the HDA, so it is visible from the outside. The stepper motor design is inferior. It suffers from slow access rate and is very sensitive to temperature. It is also sensitive to physical orientation and can't automatically park the heads in a safe zone. Besides, the actuator operates blindly from the track positions, governed only by the stop positions. Over time, the drive becomes misaligned, requiring occasional re-formats to realign the sector data with the heads.

 

The servo motor actuator is found in all modern drives, including any over 100MB in capacity. Unlike the stepper design, the heads get feedback as to position, assuring proper tracks are read. The guidance system used by the heads is called a servo. Its job is to position the head over the correct cylinder. It does this through the use of grey code. Grey code is a special binary number system in which any two adjacent numbers provide info to the servo as to their position on the drive. Also, the heads are free to move wherever needed...no steps. Basically, when the drive needs to retrieve certain data, the servo motor moves the heads out to the appropriate position on the disk and then waits for the corrects bit of data to spin over to it. The time it takes for all this to happen is called latency, and is a key measure of the speed of the drive.

 

When the hard drive is powered down, the springs attached to the heads pull the head into the platter. This is called a landing. =) Every drive is designed to handle thousands of takeoffs and landings, but since the head actually hits the platter, its best to have this happens on a section of platter where there is no data. In a voice coil design, small springs drag the heads into a park and lock position before the drive even stops spinning. This assures that the heads are not just let go of and left to drag along the platter until the platter stops, a problem common to the stepper motor design. When powered on, the drive automatically unparks itself and the parking springs are overcome by the magnetic force.

 

The spindle motor is responsible for spinning the platters. These devices must be precisely controlled and quiet. They are set to spin the platters at a set rate, ranging from 5400 RPM to 10000 RPM. The motor is attached to a feedback loop to make sure it spins at exactly the speed it is supposed to. The speed is not adjustable during operation. Some spindle motors are on the bottom of the drive, below the HDA, while the more modern ones are built into the hub of rotation of the platters, thereby taking up no vertical space and allowing more platters.

 

Attached to the spindle motor is a ground strap which helps rid the drive of the static charges created by the rotating the platters through the air. In many drives, this can be accessed by removing the logic board. After a while, this strap can become worn and produce noise, like a high-pitched squeal. One can usually lubricate the strap and stop the noise, but this entails some minor disassembling of the drive.

 

The logic board is the board of chips underneath the drive. It controls the spindle and head actuator and also translates data to a form usable by the controller and the rest of the system. Some logic boards have an integrated controller, also. Sometimes, an apparent disk failure is actually a failure of the logic board. In such a case, you can replace the logic board and regain access to the data held up on the drive. This is relatively easy to do, because the board is simply plugged into the drive and held in by screws.

 

Hard drives are precision instruments and operate mechanically, so it needs to be handled with care. They don't handle shocks very well, and you don't want to replace them mainly because its a pain in the butt. Laptop computer hard drives are still fragile, but built to handle more shock. Putting one down on a table won't hurt it, but dropping it will.

 

Capacity

Just take a trip to the computer store and you can see that there have been major advancements in hard drive technology that lead to larger capacity drives. Where there were once 500MB drives, we now have huge drives. Recently, a 36GB IDE drive was released. What led to this?

 

Well, the first thought would be: Add more platters, or maybe bigger ones. Well, yeah, larger platters would do the trick. 5.25" platters have been used on older drives, and do hold more data. But, manufacturers don't use these big platters because of the extra stresses the larger platters put on the motor. These stresses, and the simple fact that the heads have more disk to cover, make the drives hotter and a lot bigger. Most drives in use today use 3.5" platters, and 2.5" or smaller is commonly used for notebook systems. So, the manufacturer decides it would be better to keep smaller platters, but just add more of them. This works, but in order to reach such high capacities, you'd need a lot of platters. The vertical heights of these drives would just be too much. So, then what?

 

Keyword: Areal Density. This is the closeness of data bits on the hard disk to each other. Manufacturers have made major strides in this area: making more sensitive media and making read/write heads that can pack data bits much closer together. The heads operate much closer to the platter itself (but don't touch) and use a weaker electrical charge than usual to do the job.

 

This density increases performance and allows more data to be packed on a platter. The closeness of the bits together mean that more data passes the head at a time, increasing read/write performance.

 

So, now you know how they get these little drives up to 36GB and beyond.

File Structure

http://www.pcmech.com/show/harddrive/67/

 

On this page, I'll cover file structure on a hard drive. Tracks, sectors, cylinders, etc. Plus I'll cover what happens when you format and partition a drive.

 

Basically, tracks, sectors, and cylinders are the divisions of the hard drive platters where information is stored. A track is a concentric ring around the platter containing information. Since a hard drive typically has two or more platters, each storing data on both sides, these tracks line up on each platter. The identically positioned tracks on each platter are called cylinders. To better help you understand a track and cylinder, let's take a target used for target practice. You have a bunch of concentric circles, each bigger than the other, all sharing the same center, which is the bulleye. Now, each of the spaces between circles is similar to a track on a hard disk platter. Now, if you stack several of these targets on top of each other, each exactly the same, you can form a cylinder by simply taking a track and moving it down through all of the same tracks on the targets below.

Since typical hard drives are too large to deal with by the track, each track is divided into sectors. Its not that a track could not be dealt with, but since a track can hold as much as 50K sometimes, this would not be practical for storing large files. So, sectors are basically slices of the track. Different drives have different numbers of sectors per track.

 

Each sector is given an identity during formatting to aid the controller in finding what it needs in the appropriate sector. These sector numbers are written to the beginning and the end of each sector, called the prefix portion and the suffix portion respectively. These identities take actual space on the hard drive. This explains why there is a difference between the capacity of an unformatted disk and a formatted one. On a floppy, the disk itself can hold 2M or so of data. When formatted and the identities placed, the capacity reduces to 1.44M. The same holds true for a hard drive. Drive manufacturers know this and publish formatted capacities to indicate drive size.

 

There are two types of disk formatting: low-level and high-level. These both are done in the preparation of a hard drive for use. First, one low-level formats, then partitions, then high-level formats. A low-level format turns the platter from a blank slate to a divided slate. It defines the data areas: creates tracks, separates into sectors, and writes the ID numbers to each sector.

 

Partitioning segments the drive into separate areas, each capable of running its own operating system. At this point, the file allocation tables (FATs) are dropped in. There are four types of file systems.

• FAT16. The is the file system used by DOS, Win95 and Win98, unless you convert. It supports an eight letter file name max, with a three letter extension under DOS. With Windows it supports up to 255 characters. With this system, a partition can be no larger than 2G.
• FAT32. The 32 means 32-bit. This is an optional file system introduced with Windows 95 OSR-2. The file allocation units are stored as 32-bit numbers. The main advantage is that it allows for partitions of up to 2048G with smaller clusters. There is an option to convert FAT16 to FAT32 under Windows 98, and Windows NT5 will support it as an option.
• High Performance File System (HPFS). An unpopular file system only used with OS/2 or early versions of NT. File names can be 256 characters and partitions can be 8G.
• Windows NT File System (NTFS). Just like the HPFS, but only for NT.

 

FAT file system is most used in PC's today. The main problem with the original FAT was the inefficient use of disk space in defining clusters, or groups of sectors. The clusters were rather large, causing wasted space, because a small file would take up the entire cluster even though it could hold more. With FAT-32, 4 billion clusters are allowed, allowing 4K clusters. This significantly reduces disk waste.

 

The concept of FAT explains why one can run out of disk space even when you are not storing the capacity of files. For example, a 1G hard drive can run out of space with 160MB to spare. This is due simply to the FAT structure. With the original FAT, each cluster could hold 32K. But, if you are storing an 8K file, it still takes up a complete cluster, leaving the other 24K to waste. This is called slack. The only way around this is to re-partition the hard drive to two or more partitions. As the partition gets smaller the wasted space gets less. This, then is a tradeoff. The convenience of one partition, or the wasted disk space.

 

With FAT32, the wasted space is much less. Smaller clusters.

 

No matter what file system is used, a boot sector is written to the beginning of each disk, in the first sector. This sector contains the boot program which tells the system what to do when you turn it on.

 

In a high-level format, the operating system creates a structure needed to manage its files and data. In short, it creates a table of contents for the disk. While the low-level format gives a structure, the high-level format makes it readable and orderly. The DOS FORMAT command is only capable of high-level formatting on a hard disk. Most manufacturers sell drives already low-level formatted. Otherwise, a special utility is needed, usually provided by the manufacturer.

 

Hard Drive, Heal Thyself

http://www.pcmech.com/show/harddrive/158/

 

With hard drives getting bigger and bigger, and data becoming more and more volatile, as well as more and more important, having a drive that isn't going to fail is very beneficial. It is for this reason that data storage manufacturers have worked on developing a method by which the hard drive can make rudimentary attempts to diagnose its own problems, thus averting a disaster before it happens.

 

The need for drives that can do this spawned the development of the S.M.A.R.T. system. That's an acronym for Self-Monitoring, Analysis, and Reporting Technology. It was developed by a number of computer companies in a concerted effort to increase the reliability of drives. Quantum pioneered the effort in the field of hard drives, though many companies have been involved since.

 

The SMART system does just what its name implies it does. It monitors the drive for anything that might seem out of the ordinary, documents it, and analyzes the data. If it sees something that indicates a problem, it is capable of notifying the user (or, if applicable, system administrator).

 

In essence, SMART is merely a set of software tools on the drive itself, constantly running diagnostics. They run diagnostics on the motors, the media, the electronic components, and the mechanical components. Another set of monitoring software is often set up on the controller, to monitor the overall reliability of the drive, taking the data given it by the on drive software and checking it against predefined thresholds.

 

The errors that the system can detect can be predicted by a number of methods. Currently the SMART system can detect around 70% of all hard drive errors.

 

For example, motor and/or bearing failure can be predicted by an increase in the drive spin-up time and the number of retries it takes to succeed in spinning up the drive. Or, if the drive notes that the error correction is being used excessively, it can attribute this to a broken drive head or contamination, and alert before the problem gets worse. Granted, there are some things that cannot be predicted with any accuracy. An example of such would be a total electronics failure. There is no reliable manner in which to predict such a failure without highly specialized and expensive equipment, making it less cost effective. This detection of problems is not limited to only physical aspects of the drive. If it detects that the number of write errors is excessive, it can predict an increase in bad sectors and warn the user that the data should be backed up as soon as possible.

 

Most assuredly, the SMART system is a nice thing to have on a drive, despite the very high reliability of today's drives. When purchasing a drive, take this into consideration, but not too heavily. It can, yes, predict and warn against many types of problems with the hard drive. However, you must also realize that a good number of problems with hard drives (like crashes and damage through shock) are caused by users or at least are non-hardware related problems.

 

If you do want a drive with SMART protection, many companies offer them. In fact, most drives today are SMART compliant. Quantum, Western Digital, Maxtor, most good name brand drives have some level of SMART compliance.

Installing A Hard Drive - Step by Step

http://www.pcmech.com/show/harddrive/43/

 

Installing a hard drive is a medium level job. You can do it, but it might not be fun. If you are confident in yourself and would like to save the money a computer guy would charge to do it, go ahead and do it yourself. It won't be that bad.

 

Before starting, make sure you have a system disk. This is a disk that has the necessary files for your computer to boot off of. You need to make sure your system disk works now. You will need to boot your system with it in order to complete the set up of your new hard drive.

 

If you are adding a second hard drive, you need to decide which one will be the master and which one will be the slave. The master is your drive C. The other one is the slave. Look at the instructions for the hard drive. It will tell you how to make the drive a master or a slave. They usually come configured as a master, and you simply adjust a jumper on the back of the drive to make it a slave. Pre-286 computers can't handle two hard drives. A later computer can handle two IDE hard drives per IDE channel. This is more than enough for most people.

 

Get the setup in your mind. Which IDE channel? Master or slave? As a consideration, don't put the hard drive on the same channel as your CD-ROM unless you have to.

 

Physical Installation

Okay, now lets do it. If you are only installing a second hard drive or a new one, you can skip down to step 5, although this might help as a reference.

 

1. Back up your old hard drive, turn the computer off, unplug it, and take the case off. You'll want to make sure you back up your old drive first. You can do this with a tape-backup drive or some other form of removable storage. I'd recommend the later of these options, due to their speed and capacity. Also, you may want to make some quick sketches of just how everything is in there: Which direction is everything facing? Where and how are the cables connected? For some people, such sketches help to put everything back when you are done.
2. Remove the cables from the old drive. You will see both a wide IDE ribbon cable and a small 4-pin power plug. Do not force them out. The ribbon cable is usually quite easy to remove. Sometimes, though, the power connector can become stuck. Just rock it back and forth, taking care not to rip the connector off the drive.
3. Remove the mounting screws that hold the drive to the case frame. Sometimes, you may need to tip the case or get into some strange positions to reach all the screws. But, that's part of the fun.
4. Remove the old drive from the case. Be sure not to bump anything too hard on the way out.
5. Slide the new drive in right where the other one came out. If it is smaller than the drive bay ( if you are installing a 3.5" drive into a 5.25" drive bay ), you may need to add rails or a mounting bracket to make it fit. If you are adding a second drive, just pick any empty drive bay. Screw the drive into place.
6. If you need a separate controller card, install it now into any unused motherboard slot. If you are replacing a non-IDE drive with an IDE drive, you'll need to throw a new IDE controller card in. Most of today's motherboards have two built-on IDE controllers. It is easiest to use these controllers when available, and it saves a slot for something more fun.
7. Attach the cables to the hard drive. Just like a floppy drive, connect the ribbon cable and the power cable. The ribbon cable goes from the controller to the drive. Make sure the red edge of the ribbon cable is in line with Pin 1 on the drive. If you place the cable on backwards, you may get strange errors that make your new drive sound like it has died already. If you are adding a second drive, simply choose a connector on the same ribbon cable that is not used. Most ribbon cables come with three connectors: one on the end and one mid-way, then one further away on the other end which connects to the motherboard. In this case, it doesn't matter which plug goes in what drive. The computer looks at the master/slave jumpers to see which one is Master. Make the second hard drive the slave. The manual should show you how to do this on your particular drive, although many drives have the jumper settings conveniently labeled on the drive itself.
8. If you have not yet done so, replace the screws. First double check your work, though. Also, make sure you use screws short enough not to damage the drive when tightened. Do not force the screws to tighten.
9. Plug the system in , and turn it on with your system disk in Drive A:. It is best to leave the case cover off for now in case you need to fiddle with something or troubleshoot the installation.
10. New hard drives need to be prepared before they will work. You will need to configure it and set the CMOS. When you turn the system on, immediately hit the Hot Key sequence necessary to enter CMOS setup. A lot of times, this is Delete. Go to the section on IDE auto-detection, if your BIOS has this option. Follow the prompt under this section and it will auto-detect the drive. If your BIOS does not support this, then you will need to manually plug the necessary information into setup for the drive. When this is done, exit CMOS and save your changes. The system will reboot. Leave the system disk in Drive A:.
11. When the system completes boot up, it should stop at the A: prompt. Type "fdisk" and hit enter. Follow the prompts to partition the drive.
12. When FDISK is done, you should be able to switch to the C: drive, or whatever letter the new drive happens to be. Now, all you need to do is format the drive. At the A: prompt, type "format x: /s". Replace "x" with the letter of this new drive. This will proceed to format the drive and copy necessary system files to it. After that, you will be able to boot the system off the new hard drive.
13. Now you can copy files to it or whatever. If this is to be your main drive, you can install your operating system now.

 

Physical Installation - SCSI Drives

1. Take the case off. Find the drive that's in there, if you have one in there. Inspect the ribbon cables. The red edge of the cable should be facing Pin #1 on the drive. Take any notes you may need to help you install the new drive and get the cables right. Then disconnect the ribbon cable and the power supply wires. It may be necessary to remove some other parts to get at the drive.
2. Remove the old drive. It comes out the same way an IDE drive does.
3. Now you need to set any switches or jumpers that need setting on the new drive. In SCSI setups, each device gets its own SCSI ID, numbered 1-7. #7 is usually given to the adapter card. You may pick, then, any other unused address. You may need to take into account any little quirks in your adapter, such as special likings to other addresses that could cause problems a little later. You'll need the manual for this one.
4. Check for the correct termination. In SCSI setups, the adapter can hold up to seven SCSI devices. These devices are hooked up in a chain, usually with the adapter at one end and another device at the other end. This ending device must be set to be the terminating device, therefore ending the SCSI chain. Usually, SCSI devices come with a terminator plug. In some cases, the adapter is in the middle of the chain, therefore you must terminate at both ends of the chain. You may need to consult the manual for any special termination techniques particular to your brand of drive.
5. Slide the drive in and connect the cables. This is the reverse of what you did in steps 1 and 2. Make sure that pin #1 on the ribbon matches up with pin #1 on the drive.
6. Plug the system in and turn it on with a system disk in Drive A:. Continue as normal.

HDD Configuration

http://www.pcmech.com/show/harddrive/66/

 

After doing the physical installation, this doesn't mean your hard drive is ready to use. Your first step in making your system use the new hard drive is to configure the computer to it. This is mostly done in the BIOS. Most IDE drives install basically the same way. With SCSI drives, configuration procedure varies widely due to the myriad of SCSI host adapters. You must follow the instructions that come with your adapter and hard drive.

 

Automatic Drive Setup

On almost every system, the BIOS is capable of performing a special ID command on the drive. The drive sends its information to the BIOS, and the BIOS can automatically configure itself for the drive. This is very convenient, as it keeps the user from having to figure out and manually type in all the settings, although all BIOS versions allow this option as well for the brave soul.

 

To use this BIOS feature, boot the system and immediately enter the hot key sequence necessary to get into your BIOS. When there, choose the menu option for IDE auto-detection. Follow the prompts. Sometimes, it will offer you a few choices to choose from. Just choose whichever one the system recommends.

 

Manual Drive Setup

If your BIOS does not support auto-detection, or if, for some god-awful reason, you want to do it yourself, you can manually set all your drive's information. All information you need is in the manual.

 

You will need several pieces of information:

• Number of Cylinders
• Number of Heads
• Sectors Per Track
• Defective Tracks

 

After you have this info, you need to match these parameters with one of the entries in your motherboard's ROM. If there is no entry even close to yours, select "User Defined" and plug in the settings yourself. Make sure you maintain a record of the settings. If your battery dies, or your BIOS becomes corrupted, all these settings will be lossed, requiring you to retype them to get access to your drive.

 

Intelligent IDE drives can adapt to the geometry you plug in, as long as it is less than or equal to the drive's actual capacity. They translate themselves for older BIOS. This is very convenient and gets around older hard drive table entries in older BIOS.

Partitioning

http://www.pcmech.com/show/harddrive/137/

 

Partitioning is one of the necessary steps to prepare a drive for use. It is the process of defining certain areas of the hard disk for the operating system to use as a volume. A volume is a section of the drive with a letter, like C: or D:. All hard drives must be partitioned, even if they will have only one partition called C:.

 

A partition program writes a master partition boot sector to cylinder 0, head 0, sector 1. The data in this sector defines the start and end locations of each of the other partitions. It also indicates which of these partitions is active, or bootable, thus telling the computer where to look for the operating system.

 

All systems can handle 24 partitions, either spread out on the same drive or many drives. This means that one can have up to 24 different hard drives, according to DOS. DOS can't recognize more than 24 partitions, although some other OSes may. The limiting factor is simply the availability of letters. All partitions must have a letter. There are 26 letters, A: and B: are reserved for floppy drives, leaving 24 letters available.

 

Although there are third party partitioning programs that boast added capabilities, DOS FDISK is the accepted program for partitioning. FDISK sets up the partition in a way optimum for DOS, and allow more than one OS to operate on one system.

 

FDISK only shows two DOS partitions, the primary partition and the extended partition. The extended partition is divided into logical DOS volumes, each being a separate partition. The minimum partition size is one megabyte, due to the fact that FDISK in DOS 4.0 or later create partitions based on numbers of MB. Partition size is usually limited to 2G. DOS versions earlier than 4.0 allow max partitions of 32MB. Using the Fat32 system under DOS 7 and Windows 95 OSR2, max partition size is kicked up to 2T, or 2,000G.

 

How To Partition

The first partition is your primary DOS partition. This is your C: drive and can't be divided. This is also called the active partition. You can only have one active partition.

The second partition is optional. It is called an extended partition. This is the space left over after the primary partition. Each extended partition must be labeled with a letter D: through Z:. In FDISK, there is one extended partition, with it being divided up into Logical DOS Drives which each have a drive letter.

 

To start this, type "fdisk" at the A> prompt. If this doesn't work, it is because your drive is not installed correctly.

 

First you have to setup a primary DOS partition. Choose Option 1 ( Create DOS partition or Logical DOS drive). Choose Option 1 in the next menu. Now you can make your entire C: drive the primary partition or only a part of it. Many people just make the entire drive one partition just to stay simple. If you want to break from this norm, specify the amount of drive you want to partition in either megabytes or percentage of total drive. If you are using a percentage, be sure to follow the number by a "%" or the computer will think you're talking MB's.

 

Next, you'll need to make this partition active. Return to the main FDISK menu and choose Option 2 ( Set Active Partition). Follow the prompts.

 

If you're going to create an extended partition, choose Option 1 again, but this time choose Option 2 in the next menu ( Create Extended DOS partition). Plug in the percentage of drive to partition for this one. Do not make this partition active. Only one can be active.

 

After you create an extended partition, you will be given the Create Logical Drives option in the extended partition menu. Follow the on-screen instructions to assign drive letters to your partitions D: through Z:. Keep in mind that D: is often used for the CD-ROM.

 

After all this is done, you can choose Option 4 ( Display Partition Information) and check your work.

 

Optional FDISK Functions

FDISK in DOS 5.0 or later is more powerful than most people know. There are several options available with the program that are undocumented in the DOS manuals. The bad news is that these command are unavailable with Windows 95. Instead, you will have to purchase a third party program such as PartitionMagic.

 

The most useful, in my opinion, is the "/MBR" parameter. This parameter tells FDISK to rewrite the Master Partition boot sector based on the partitions present on the drive, without damaging the partitions on the drive. This is very useful when recovering from a virus that infects the boot sector of the drive. Use it by typing "FDISK /MBR" at the A> prompt.

 

To back up the partition table onto a floppy diskette, type "MIRROR /PARTN". This uses the MIRROR program to copy the partition table into a file called PARTNSAV.FIL. This can then be stored on your system disk. To restore this partition info, type "UNFORMAT /PARTN".

Formatting

http://www.pcmech.com/show/harddrive/58/

 

Formatting is another necessary step to hard drive preparation, and very simple. In most cases, when installing a new hard drive, all you need to do is a high-level format. It is usually the final step in preparation.

 

When preparing a new drive, use the "FORMAT C: /S" command. This high-level formats the volume C:, copies hidden operating system files to the volume, and prompt you for a label. It marks bad sectors as unreadable, writes the boot sector, creates the FAT, writes the root directory, and copies system files. If you do not want to copy system files to the you are formatting, just type "FORMAT X:" where X is the drive letter you wish to format. You cannot format a drive while working on that drive. Meaning, you cannot be at the C prompt and try to format the C drive. You must switch to another drive which contains the FORMAT.COM file. This is the file used to format drives.

 

The other type of formatting is the low-level format. In general, this procedure is already done on your drive when you buy it. Only on old drives would this need to be done. Other situations exist where you would want to low-level format your hard drive. If you need to erase all traces of data on the disk, a low format will do this. It will also remove corrupted operating systems or viruses. It will also re-map the drive so as to reallocate all bad sectors to other sectors. This basically replaces bad sectors with good ones. It will make your drive appear to be free of defects. This process is called defect mapping.

 

That said, manufacturers recommend you never low-level format a hard drive.

 

A low-level format cannot be done with the FORMAT command. It is recommended you get a low-level format program from the manufacturer of your drive. These programs are tailored to work with specific drives and can sufficiently trace the defects and map them. Visit the web site of the manufacturer to find these programs. They are often available for download.

ST-506/412 & ESDI Interface

http://www.pcmech.com/show/harddrive/162/

 

Wow, does that title look like Greek of what?! But, really, they are interfaces used for hard drives.

 

The ST-506/412 interface was developed by Seagate Tech around 1980. The first drive to use it was, guess what, the ST-506. This was a big drive with the 5.25" form factor, very large in comparison to today's drives. A year later, Seagate came out with the ST-412. This drive added a new feature to the interface called buffered seek. Although most of these drives are long gone now, Seagate still manufactures drives for many PC's. This contributes to the popularity of the ST-506 interface, although it doesn't compare to the IDE or SCSI interfaces.

 

The ST-506 interface is very simple to use. It requires no special cables or connectors. It will work with any ST-506 controller. The only worry is the BIOS in the system: does it fully support the interface?

 

In the beginning of the interface, BIOS support on the motherboard was rare. Instead, the support came from a ROM BIOS chip on the controller itself. With the release of the AT machine, which constitutes almost every machine now days, support for the interface was written into the motherboard's BIOS. Virtually all systems in use today have ST-506 support.

 

ST-506/412 isn't much used in mainstream systems today. It is not conducive to high-performance drives, being that it was originally designed for old drives. Most drives boasting this interface are smaller than 200MB. Due to these limitations, the interface is obsolete. The only reason I mention it at all is because it is used in many older machines, like the XT's.

 

ESDI stands for Enhanced Small Drive Interface. It is another old interface developed by Maxtor in 1983. It was developed as a high-performance successor to the ST-506 interface. The interface is built to out perform the ST-506. The endec, or encoder/decoder, is built onto the drive itself. It is capable of 24 MB/sec data transfer, although most setups limit it to about 15MB/sec.

 

ESDI never took off. With several different versions out, it could never compete with the low-cost, high-performance IDE interface. I doubt you'll ever find a new PC using the ESDI interface. It was a 1980's interface.

IDE Interface

http://www.pcmech.com/show/harddrive/78/

 

Integrated Drive Electronics (IDE) is really a misnomer in the way we use it today. IDE really refers to any drive with the controller built-in. The interface most of us use, that we call IDE, is actually called ATA, or AT Attachment.

 

Most drives today are IDE. These drives have the controller built on. They plug into a bus connector on the motherboard or an adapter card. Such drives are easy to install and require a minimum number of cables. This is due to the fact that the controller is on the drive itself. Less parts are needed and the signal pathways can be much shorter. These short signal pathways improve reliability of the drive. Before, data could lose its integrity while traveling over cheap ribbon cables. Lastly, integrating the controller is easier on the manufacturer because they do not have to worry about complying with another manufacturer's controller. Each drive is an independent entity.

 

As said before, IDE is really a much broader term than what we usually use. Most of the time, one is referring to ATA IDE, simply because this is most popular. There are other types, including MCA IDE and XT IDE. These will be discussed briefly further down.

 

ATA IDE

This is the most popular IDE form. CDC, Compaq, and Western Digital were the first to create the interface. They also decided to use the 40-pin connector. They were large drives of the 5.25" form, but were only 40M. They were used in the early Compaq 386 systems, using WD controllers. Later, Compaq founded Conner. Conner produced drives for Compaq, but was later sold.

 

In the late 1980's, the ATA IDE was set as ANSI standard. This caused all manufacturer's to agree with a common design for the interface. But, before this was done, many companies had produced their own variations. This sometimes makes it hard for us to make these older drives work with newer ones in the same system. Some areas of the ATA standard were left open to manufacturer's for their own commands. Due to this, the standard is really loosely set. Low-level formatting drives, then, require a program tailored to drives from a certain manufacturer, one that knows that company's commands.

 

Dual Drives

Using two drives in the same system has been known to be hard at times. This is usually due to the fact that each drive has its own controller, both trying to operate over the same bus. One of the nice features introduced with ATA was the ability to operate two drives together in a chain. The primary drive is the master, and the second drive is the slave. On most drives, you tell it to be a master or a slave with a jumper on the drive itself.

 

When two drives are on the same ribbon cable, all commands are received by both controllers. Each drive must respond only to commands meant for itself. This is done with that jumper. Setting the drive as either master or slave tells it to ignore the commands for the other drive and to only act on ones meant for itself.

 

ATA I/O

The ATA interface uses a 40-pin connector. This is usually designed to prevent plugging it in backwards. This design is recommended. In theory, plugging it in backwards can damage the drive and related circuitry, although I have done it before a few times and all my drives still work.

 

The ribbon cable is 40-wires wide. It carries all signals to and from the controller. This cable should be no longer than 18 inches long.

ATA Types

Non-Intelligent IDE was the first type. These drives were simplistic. They only responded to the first eight commands built into the original WD1003 controller. They were actually more like ST506.412 drives with the controller screwed on. Most of these drives could be low-level formatted, unlike today's drives. Each was low-level formatted in the factory with a few optimizations built on. Factory defects were written as a file to the drive. This means that, although you can low-level format the drive, it would erase the factory optimizations and defect list. Some companies released programs to do this while saving these settings, but many did not.

 

Intelligent IDE drives were enhanced to use special commands like the "Identify Drive" command.

 

Intelligent Zoned Recording IDE is an intelligent drive with special Zoned Recording capability. This means that the drive can have a different number of sectors on each track. Since the BIOS can still only handle a fixed number of sectors per track, the drive runs in a special translation mode. This ability means that you cannot low-level format this type of drive without a special program from the manufacturer.

 

ATA-2 is EIDE, or Enhanced IDE. This is an extension off the original ATA that includes features such as PIO and DMA modes. These are basically performance enhancing features and are discussed below.

 

The main benefits of ATA-2 are:

• Increased Capacity. This is basically due to an advance in BIOS to work with drives larger than 504 MB. This limit was there basically because of the geometry in the drive. Newer enhanced BIOS are capable of using translation modes, thereby using different geometry when talking with the drive than when talking with the software. If your BIOS is dated around 1994 or later, it is probably enhanced. You can tell if it offers settings called LBA, ECHS, or Large, which are just three different methods of translation.
• Faster Data Transfer. This is what everyone is after these days. ATA-2 offers several different modes for higher performance. Most drives today are capable of PIO Modes 3 and 4, which are very fast. PIO (Programmed I/O) modes determine the speed at which data is transferred to and from the drive. Below is a table of the PIO Modes:

 

 

·         To run Mode 3 or 4, the IDE port must be on a VL-bus or PCI bus connection. Some newer boards with two IDE connectors only have the IDE 1 connected to the PCI bus, while the second IDE connector uses an ISA bus, only capable of Mode 2. One should look into this before buying a new motherboard.

• ATAPI. This is ATA Packet Interface. This is designed for extra drives like CD-ROM's and tape drives that connect to an ATA connector.
• DMA Transfer. ATA-2 drives support Direct Memory Access transfers, which means that data is transferred directly from the drive to memory, bypassing the CPU. These drives only support the feature. Most OSes do not support it. Proper software support for this transfer is needed before it can be taken full advantage of.

 

Cable Configuration

Cable configuration is quite simple with the ATA IDE interface. There is a single 40-pin cable with three connectors on it. One of these connectors plugs into the IDE connector on the motherboard or I/O adapter card. The other two attach to the drives. On most setups, one end of the cable is attached to the IDE connector. The middle connector attaches to the secondary drive, if there is one. The other end is attached to the primary drive, or drive C:.

 

There is no termination of the chain required, as there is in SCSI. A termination circuit is built into the drive.

 

Although the above is the typical setup, it isn't necessary. Sometimes, the D: drive is on the end while the C: drive is connected in the middle. This usually works fine since the master/slave relationship is determined by the jumpers, not the cable. Other setups have the middle connector attached to the motherboard, with the cable ends attached to the drive: a sort of Y arrangement. This is done in many systems, but must be handled with care because the master/slave relationship is then determined by position on the cable. On the Y setup, a special signal called the CSEL, carried on pin 28, defines primary or secondary. If the CSEL circuit is closed, the drive is primary. If it is open, the drive is secondary. This is usually done with a small hole pricked through wire 28 on the cable. Whichever drive is connected to that section of cable is then drive D:. Get it?

 

Jumper Settings

Most IDE drives come in three configurations: Single drive, master, and slave. These are controlled by a small series of jumpers, usually on the rear of the drive. The single drive setting tells the drive it is alone in the system, and it responds to all commands. If it is configured as a master, this tells the drive there is a slave drive present, and the drive will respond to only master commands. If the drive is configured as a slave, it responds only to slave commands. These jumpers are usually labeled on the drive, so setting them should be no problem.

 

Some drives also have a "Slave Present" jumper. This is only needed on the master drive, and basically just tells it that it has a partner.

 

Before the ATA IDE specification, there was no common method of master/slave relationships. Each manufacturer had a different method. For this reason, these drives can be difficult to work with in a two-drive system. Some must work in either a master/slave or slave/master order.

 

Now a quick blurb about the older IDE's.

 

XT-IDE

The XT IDE drive was used in older XT systems with the XT ISA bus. Only IBM, Seagate, and Western Digital made them, and they got no bigger than about 40M.

 

MCA IDE

Computers with the MCA bus, at one time, used a MCA IDE drive. It is only an IDE drive configured to work over the MCA bus. It is not compatible with any other type of bus. Very few companies made them, therefore they are very expensive, and small. Most systems today with the MCA bus use SCSI drives.

SCSI Interface

http://www.pcmech.com/show/harddrive/153/

 

SCSI is an entirely different interface than the more popular IDE. It is more of a system level interface, meaning that it does not only deal with disk drives. It is not a controller, like IDE, but a separate bus that is hooked to the system bus via a host adapter. A single SCSI bus can hold up to eight units, each with a different SCSI ID, ranging from 0 to 7. The host adapter takes up one ID, leaving 7 ID's for other hardware. SCSI hardware is typically hard drives, tape drives, CD-ROMs, scanners, etc.

 

SCSI's popularity is increasing. Speed seems to be the main reason for this, although I will show further down that this really isn't anything to get excited about. One advantage is that there are a multitude of hardware types that can use a SCSI bus. The interface is very expandable, whereas IDE is pretty much limited to hard drives and CD-ROMs.

 

The reason for the slow taking of SCSI is the lack of standard. Each company seems to have its own idea of how SCSI should work. While the connections themselves have been standardized, the actual driver specs used for communication have not been. The end result is that each piece of SCSI hardware has its own host adapter, and the software drivers for the device cannot work with an adapter made by someone else. So, due to the lack of an adapter standard, a standardized software interface, and a standard BIOS for hard drives attached to the SCSI adapter, SCSI is pretty much a mess for the end-user. Don't get me wrong, here, though. SCSI is a relatively easy thing to implement, should you wish to.

 

SCSI Evolution

SCSI has come a long way. In the beginning, one couldn't even use a hard drive on the bus. This was mainly because the BIOS in those systems were designed to use the ST506/412 controller. With the IDE, the BIOS was easily changed because of the similarity to ST506/412 on the WD1003 controller. At the register level, though, SCSI was very different, and would have required an entirely new set of BIOS in the PC. The newer PC BIOs has been designed for SCSI support or there is an extension BIOS on the host adapter.

 

When this feature first started, though, hard drives could only be used with DOS. Later, Adaptec and Future Domain designed adapters that could be used with non-DOS OSes.

Many high-end systems have built-in SCSI support. There is usually an adapter card or an adapter built in to the motherboard. This native support for SCSI was set in motion by IBM. Their example was followed by many manufacturers. As a result, SCSI integration is becoming very easy to work with and will get easier as technology progresses.

 

SCSI Standards

SCSI-1 was standardized by ANSI in 1986. While this outlined the physical and electrical traits of SCSI, it failed to outline a common set of commands so that all manufacturer's hardware would work together. The industry, then, decided to agree on a minimum set of 18 basic commands. This command set was called the Common Command Set (CCS). All SCSI hardware supported the CCS.

 

CCS became the basis for SCSI-2, a more advanced version of the original SCSI that provided extra commands for other types of devices. SCSI-2 also provided extra speed with options called Fast SCSI and a 16-bit version called Wide SCSI. A feature called command queuing gave the SCSI device the ability to execute command in an order that would be most efficient. This is most useful on hard drives using OSes that are multitasking.

 

The standard for SCSI-1 and SCSI-2 is somewhat clouded. Almost all features and commands of SCSI-1 are supported in SCSI-2, and most SCSI-1 hardware is called SCSI-2. Many manufacturers boast that their equipment is SCSI-2. This makes it seem better, but in reality, it may not support the extra features that were included in the true SCSI-2 revision.

 

This also means that SCSI-1 adapters will work with SCSI-2 hardware. SCSI-1 and SCSI-2 compliant hardware is the same.

 

A SCSI-3 standard is being worked on, although features of this new standard are already in use by some manufacturers. Such drives run in Fast-20 mode or an Ultra-SCSI mode. These speed rates are defined below.

 

Data Transfer Rates

Below is a quick table describing the data transfer rates:

 

Fast SCSI delivers a 10 MB/sec transfer rate. When combined with the 16-bit bus, this doubles to 20 MB/sec. This is called Fast-Wide SCSI.

 

Ultra SCSI, also called Fast-20 SCSI, is twice as fast as Fast SCSI. It is part of the SCSI-3 setup, which has not been standardized but is still being sold in high-speed drives. Ultra SCSI delivers 20MB/sec over the 8-bit bus. Ultra-Wide SCSI incorporates the 16-bit bus, and the speed raises to 40MB/sec.

 

SCSI-2

Now that I have jarbled the subject up saying that SCSI-1 is the same as SCSI-2, let me say that SCSI-2 is improved. SCSI-1 and SCSI-2 work the same: just one has more features than the other. SCSI-2 hardware will work over a SCSI-1 adapter just fine, but the extra features won't be able to be used.

 

What are the extra features?

• Termination. SCSI-1 was very stringent in its requirements in termination. It used a 132ohm passive terminator. This was not conducive to the higher speed data transfers used today and sometimes caused data errors when more than one device was added to the chain. SCSI-2 uses an active terminator, or voltage-regulated. It lowers the impedance of the termination and improves reliability.
• New Commands. SCSI-2 made the CCS part of the standard. It also rewrote some old commands and added several new ones. These new commands were designed for hardware other than hard drives.
• Command Queuing. While SCSI-1 adapters could only send one command at a time, SCSI-2 adapters could send 256 to one device. The device, in turn, could store the commands, process them, and reorder them separate from the bus.

 

SCSI-3

As said above, elements of SCSI-3 are in use today in the forms of Ultra-Wide and Ultra SCSI drives, but the SCSI-3 standard has not yet been agreed upon.

 

There are many interesting advances with SCSI-3. For example, while SCSI-2 can support up to 8 devices on a single chain, SCSI-3 will support 32.

 

SCSI-3 also hold promising developments such as Serial SCSI. This feature will allow data transfer up to 100MB/sec through a six-conductor coaxial cable. It will solve many of the termination and delay problems of older SCSI versions. It may also ease SCSI installation woes by being more plug-and-play in nature, such as automatic SCSI ID assigning and termination.

 

Termination

The SCSI bus operates on a chain, and like all other interfaces, it must be properly terminated at the end of the chain. There are three types of terminating devices:

• Passive - This is basically a bunch of resistors. It does the job, but higher performance drives can't use it and it only works on distances of 2 or 3 feet.
• Active - This terminator has a voltage regulator that ensure the correct termination voltage. Many have a LED showing the termination level.
• Forced Perfect Termination (FPT) - Even better than Active termination.

 

Configuration

SCSI drives aren't that hard to configure.

 

Each device must have a SCSI ID, 0-7. The host adapter takes one ID. Most are usually factory-set to ID 7, which is the highest-priority ID. Many adapters require that any SCSI boot drive be configured to a certain ID. With the newer ones, it doesn't usually matter.

 

The ID is configured by some type of switch or jumper on the drive, much like the master-slave jumper on an IDE setup. There are three jumpers used to describe the SCSI ID. Instead of making this simple, manufacturers decided to make the ID # a result of a binary representation of the jumpers. For example, setting all three jumpers off gives a binary of 000, meaning SCSI ID 0. Below is a table of jumper settings:

 

Depending on the manufacturer, the order of these jumpers may have been reversed. In this case, just flip the order of the jumper settings around. For example, ID 4 above is on-off-off. On a reversed setup, it would be off-off-on.

 

Besides configuring the proper ID, proper termination must be ensured. If the adapter is at the end of the chain, enable its termination. If it is in the middle, disable its termination and install termination at each end of the bus. Use the best-terminators possible. Passive are bad, Active is better, but FPT is best. Stick with high-quality terminators and you will avoid most termination problems.

 

There are a few other settings available:

• Start-on Delay - Controlled by jumper. This delays start up of the drive. When you turn on the system, you can overload the power supply if all drives try to start spinning at the same time. To avoid this, enable this setting. This keeps the drive from starting until it gets a Start command1. Most adapters send this command in succession, from 7 down to 0.
• SCSI Parity. This ensures data integrity. It should be enabled on every device. It is only an option because older adapters didn't support it.

 

Tips

• Add one device at a time.
• Use good termination. Use the best available terminator.
• Document your settings. All of them.
• Use high-quality cables. The cables should be rated for SCSI and all be the same make. Also, don't use more cable length than the bus allows.

SCA SCSI Drives

http://www.pcmech.com/show/harddrive/152/

 

UDMA, ATA, EIDE, and SCSI, buzz words that we all associate with hard drives today but in the past 8 months another word to pop up is SCA (Single Connection Attach) SCSI. It seems even with all the "gurus" out there that say and know a lot about computers that word still seems to confuse most. I am sure most experienced shoppers have seen these types of ads in the past 8 months:

 

"SCSI SCA Fast Wide Seagate Hawk 15230WC - $199.00 (adapter sold separately)"

No it’s not a hoax and not a scam but rather a different type of SCSI that very few people know about. In this article you will have a little better knowledge on SCSI SCA drives so you will know what someone is talking about when they try to sell you a hard drive for your SCSI system.

 

First of all SCA drives are a bit different than the regular SCSI drives. In the chart listed below you will see the major differences between the SCSI SCA and regular SCSI drives:

 

	Seagate Hawk SCA15230 WC	Seagate Barracuda UWST34371W
Interface	80 pin SCA	68 pin
On board Termination?	NO	YES
On board SCSI ID’s	NO	YES
Powered by 4-pin Molex?	NO	YES
PROS	·         Cheap storage. Enough Suppliers out there that are willing to make deals.	Can be used in a single drive system without external termination needed. Can be bought in OEM and Non-OEM. Fast!
CONS	Cannot be used in a single drive system. Needs to be middle drive or have an attached terminator. Also you must have an SCA to 50 or 68 Pin Adapter that retails for $25.00. Reliability unknown and cannot be bought Non-OEM. Slower than Non-SCA drives because of the use of adapter.	Higher price, 68 pin internal and external SCSI cables can be expensive.

 

I myself started on my quest for SCSI Zen by buying a 4.3GB Fast Wide Seagate Hawk thinking that I would like to get into the world of SCSI. Because of my love for graphics I figured the extra cost would be a great investment and that my knowledge set would help me setup the SCSI system with ease.

 

Wrong! First of all after buying a new Diamond Fireport 40 UW SCSI card I found that I couldn’t get the drive to work unless I bought a separate $25.00 adapter. So investing more money I bought one, they shipped it and voila into the system it went. The particular card I bought would handle 50 and 68 pin hookups so I opted for the 68 pin route, jumpered the "term" pins for termination and plugged it all in. Thinking I was near the end of my IDE life I turned on the computer booted up and started working with the usual Adobe PhotoShop 25mb files. Quick you say? Hmmm not so quick. As a matter of fact the hard drive would seem to stall in the middle of a job like umm opening a window or just minor things. This is supposed to be SCSI I thought but then decided to make some phone calls.

 

First of all I called the vendor who sold me the drive. They stated that the drive had no termination on it but the adapter they sold me would work and terminate the end of my SCSI chain. They stated that they have sold the drives without any problems and that the drive should work properly.

 

Still skeptical I called Seagate Technical Support directly and talked to a gentleman who proceeded to inform me of the drive that I had just purchased. He stated that:

 

1.      Seagate does not support this drive because that it is an OEM drive that was made for computer companies such as Hewlett-Packard for their servers and that the drive was intended to have a backplate that controlled the power, termination and ID’s.

2.      That the drive I had bought wasn’t recommended on any home user's computer and that the only place where I could possibly get a good adapter would be from http://www.scsi-cables.com.

 

So I called CS Electronics and talked to a representative that said that for me the drive could not terminate without their adapter and their adapter only. It seems at the time that they didn’t even offer a 68-pin adapter that would provide active termination. They also said the best thing to do is to buy a 68 pin internal SCSI cable and to attach the middle to the drive and then put an active terminator on the end and the opposite end attach it to the card. They said that it would complete the SCSI chain and I wouldn’t have any more problems with the drive. Hmmm another $60.00 dollar I would have to spend…..

 

So in the end I am sure you are asking what I did to solve my problem I assume…..well I went out and bought a regular 4.3 GB SCSI UW Barracuda and put that on the end and am currently using the drive for storage only. It’s been almost 8 months and I haven’t had one problem using the Hawk as a storage and storage only. I suppose it has served it’s purpose even if it has caused me trouble since unpacking it from UPS but it also has given me the knowledge on SCA drives to share with others.

 

Just a quick note: I heard a rumor that HP ordered 10,000 of these drives for their servers and then backed out of the deal, forcing Seagate to sell these drives as OEM to wholesalers……don’t quote me but that is the rumor :o)

 

SCA Drive vs. SCSI Non SCA Benchmarks

After owning both SCA and non-SCA drives I did some benchmarks to determine if the SCA Drive with an adapter was any slower than a non-SCA drive even if they were the same brand and model. Here are the results:

 

 

As you can see the Seagate Hawk reported slow scores due to the low RPM and also due to the SCA adapter. As for the SCA Barracuda it did better because of higher RPM yet still fell short of the non-SCA Barracuda that spins at the same rate and is built exactly the same minus some features that are disabled at the factory. The Seagate Barracuda non-SCA reported fast scores and also allowed the enabling of write cache, which improved the overall speed of this drive.

 

Also you can note that both the SCA drives show up as "Sun 4.2" in the bios when you boot into windows and also when you use a hard drive utility. This clearly shows this is an OEM drive made specifically for a certain manufacturer and that it was not intended for public use. A representative at CS Electronics stated that by removing the embedded tag would possibly improve the performance but he could not tell me how to do it and was quite debatable whether that would affect performance by Seagate.

 

Yet with all these hard drive scores, none are downright horrible, they just all have different uses. For a primary drive to run your Operating System off of I would choose a non-SCA drive like the Seagate Barracuda. As for storage I would and I do use a SCA drive if the price is right and if there is a warranty provided by the seller. So play it smart, get educated and good luck and happy shopping!

 

System Configuration:

Intel Pentium II 233

Abit LX6 Motherboard

80mbs SDRAM with EEPROM (Samsung)

Diamond Fireport 40 UW SCSI Card

Windows NT 4.0 with Service Pack 3 installed

IDE vs. SCSI

http://www.pcmech.com/show/harddrive/79/

 

The popularity of SCSI is increasing rapidly, but I believe this is due to a misunderstanding. It is often thought that SCSI automatically blows IDE away when it comes to performance. While SCSI does offer a faster throughput, one's activities on the machine affect just how much this performance will really matter. Several factors must be considered when determining which is better for you.

 

Performance

Most PC's use IDE drives because they are cheap and they perform well. But, to look at performance, you need to look at the entire drive.

 

Many manufacturer release identical model drives in both IDE and SCSI formats. If you look at these drives, they are identical except for the logic board. this means that the HDA and other drive mechanics are the same. The difference lies in the logic board. The IDE logic board has the disk controller and the built on AT bus interface. The logic board on the SCSI drive contains one extra SBIC chip. Basically, this chip is a SCSI adapter to allow the drive to operate on a SCSI bus. So, structurally, IDE and SCSI drives are the same.

 

The performance overhead of SCSI over IDE comes from structure of the bus, not the drive. The nature of the SCSI bus allows it much better performance when doing data hungry tasks such as multi-tasking. The SCSI bus controller is capable of controlling the drives without any work by the processor. Also, all drives on a SCSI chain are cable of operating at the same time. With IDE, one is limited to two drives in a chain, and these drives cannot work at the same time. In essence, they must "take turns".

 

Comparison

In some computers, SCSI is better. As mentioned above, SCSI is a smarter bus than IDE. There are many steps in the SCSI data transfer. But, on OSes that allow multitasking, or if you often use several programs at once, the SCSI drive is a better choice because this extra intelligence of the SCSI bus is used.

 

SCSI devices can communicate independently from the CPU over the SCSI bus. This is due to the fact that each device has its own embedded controller. Data can then be transferred at high-speeds between the devices without taking any CPU power. IDE, likewise, uses controllers on each device, but they cannot operate at the same time and they do not support command queuing.

 

Last Thoughts

Finally, let me say that for most people, IDE is just fine and offers very good performance. The reason I believe one does not need to get SCSI, though, is that most users do not use their system in a way that would actually justify the SCSI bus. While the nature of the bus is faster, it takes certain situations to actually need it. Couple this with the significantly higher price, one can see that they can easily live with IDE.

Ultra ATA/66

http://www.pcmech.com/show/harddrive/179/

 

As always, things get faster and better. It's how humanity has lived ever since the Greek's introduced the Olympics. With computers, it's the same thing as 100 meter sprinters. Every year, things get faster and therefore better.

 

Ultra ATA/66 utilizes an 80 wire, 40 pin IDE cable, which can be thought of as a regular IDE cable, with extra wires leading no where to prevent line noise. The name insinuates that the hard drives constantly runs at 66MB/s, that's only the maximum burst transfer rate of the interface between the hard drive and the host. The reason that this cannot be the sustained transfer rate is because the hard drive, being mechanical and not electrical in nature, is held back by physical constraints.

 

The hard drive is just a series of disks built inside a housing that spin at a certain speed. From a technical standpoint, it appears to be the same as most other drives. Most today spin at 5400RPM, while some spin at 7200RPM. For instance, if the Hard drive fills it's buffer, then a few seconds later, the CPU requests for some data, which just so happens to be in the buffer, to be sent to the main system memory. Then, all the data in the buffer will be sent at 66MB/sec. The rest of the data needed, if there is any, will be read off the hard drive's platters, then sent to the buffer, where it would be sent at 66MB/sec to the system memory. This is where the physical constraints come in. The hard drive cannot read things at 66MB/sec. Currently, the fastest drive on the market can read at approximately 56MB/sec. This keeps us 10MB/sec shy of that maximum rate, and even the 56MB/sec rate is under conditions so perfect no computer user can hope to achieve.

 

Now, as for the 80 wire cable. It's a standard IDE cable, with 40 holes in each connector. That means it's backwards compatible with older Ultra ATA/33 and EIDE hard drives. The reason for the extra 40 wires is to reduce the problem of signal interference. At 66MB/sec, the electrical signals are sent so fast that they sometimes get crossed in the small space between them. The extra cables, in the Ultra ATA/66 suck up those portions signals that get away and take them nowhere. They basically just eat them and the signals disappear. That way, the cable can keep running at 66MB/sec, with no problems.

 

Ultra ATA/66 was introduced early. That was for a reason. That way, it gives companies time to catch up and perfect Ultra ATA/66 by the time it is needed. I recommend holding off purchasing a Ultra ATA/66 hard drive or controller until hard drives get faster. As for now, unless you have one of those incredibly fast drives, its not going to be necessary to have ATA/66. Stick with the ATA/33 for a while.

ATA/66 vs. ATA/33

http://www.pcmech.com/show/harddrive/6/

 

In the dark ages, back in May, Jeff used an old Asus motherboard. The day after the dark ages, Jeff had switched to a new Abit motherboard, with bells, whistles, gadgets, and even a doohickey. To better play with the motherboard, Jeff purchased a 20GB hard drive to replace his measly 6.4GB hard drive. He transferred all of his vital information to the drive, and sold the drive that he replaced on eBay. Soon afterwards, while bragging to Dok, he mentioned that the drive was ATA66 and that he has an ATA66 controller. And so begins our saga…

 

And forth went Jeff to write a comparison review article between ATA33 and ATA66, which he would be able to do with his magnificent motherboard. But alas, Dok is a bastard and wanted Jeff to reformat his drive, which has all of his crap on it. For generations (read weeks), Jeff and Dok were at each other’s respective throats, Dok wanting the drive reformatted, and Jeff wanting all his data to not disappear into oblivion. Cries for blood echoed in Dok’s camp, while cries for sleep echoed in Jeff’s. But one side had to finally give.

 

Now, with much pomp and ceremony, we present to you the results of the comparison review. With great skill and intestinal fortitude Dok called upon the great people at Quantum, who bailed a drive on us for review.

 

Is ATA/66 really necessary.  We've seen all of the synthetic benchmarks saying that ATA/66 is twice as fast as ATA/33, which it should be, being it transfers data at twice the speed.  In the real world, it doesn't transfer twice as fast.  That is mainly because the interface isn't the bottleneck in today's drives.  It's actually the rotational speed of the disks inside the hard drive.  The majority of the drives on the market today, be them ATA/33 or ATA/66, are spinning at 5400RPM.  Some of the newer and faster drives are spinning at 7200RPM.  Even with the 7200RPM drives, as we are about to prove, the real world difference between the two interfaces isn't noticeable, because the rotational speed is still a bottle neck.

 

By taking the same hard drive, with the same rotational speed, and cache site, and placing it on the ATA/66 and ATA/33 Channels of the same motherboard, in the same system.  The only thing that is different is the ATA Channels, even the operating system is the same.  Below are the Synthetic Benchmarks is WinBench 99 Version 1.1

 

Test	ATA/66	ATA/33	Difference
Business Disk	5900 Thousand Bytes/Sec	4980 Thousand Bytes/Sec	18.47
High-End Disk	17100 Thousand Bytes/Sec	12400 Thousand Bytes/Sec	37.9
Disk Transfer Rate Beginning	27000 Thousand Bytes/Sec	8620 Thousand Bytes/Sec	213.22
Disk Transfer Rate End	19600 Thousand Bytes/Sec	8490 Thousand Bytes/Sec	130.86
Access Time	11.7 Milliseconds	12.2 Milliseconds	4.1
CPU Utilization	3.87 Percent Used	49 Percent used	Ungodly Huge
Playback Business Overall	5900 Thousand Bytes/Sec	4980  Thousand Bytes/Sec	18.47
Playback High End Overall	17100 Thousand Bytes/Sec	12400 Thousand Bytes/Sec	37.9
AVS/Express 3.4	12400 Thousand Bytes/Sec	8490 Thousand Bytes/Sec	46.1
Front Page 98	64600 Thousand Bytes/Sec	64700 Thousand Bytes/Sec	-.15
Microstation SE	18700 Thousand Bytes/Sec	18600 Thousand Bytes/Sec	.53
Photoshop 4.0	9760 Thousand Bytes/Sec	6450 Thousand Bytes/Sec	51.3
Premiere 4.2	15200 Thousand Bytes/Sec	11100 Thousand Bytes/Sec	36.94
Sound Forge 4.0	2900 Thousand Bytes/Sec	16800 Thousand Bytes/Sec	-82.73
Visual C++ 5.0	17900 Thousand Bytes/Sec	13800 Thousand Bytes/Sec	29.71
Disk Playback/ emovable Media	6930 Thousand Bytes/Sec	5050 Thousand Bytes/Sec	37.22

 

As expected, ATA/66 isn't 100% faster than ATA/33, even in synthetic benchmarks.  Only in the Disk Transfer Rate tests does ATA/66 perform more than 100% better than ATA/33, which is probably a fluke in the benchmarking program, or the system that we used.

·         Pentium iii Katmai 500MHz Slot 1 SECC2 Processor

·         196MB of PC100 SDRAM

·         Abit BE6 Motherboard

·         10.2GB Quantum Fireball Plus LM Ultra ATA/66 with 7200RPM Rotational Speed

·         Voodoo3 2000 AGP

·         Etherlink III PCI, DLink Ethernet PCI Network

·         Soundblaster 512 PCI

 

Enough with the Synthetic Benchmarks, those tell you hardly anything about the real world.  It's the actual performance when you're using it.  And what can be more trying on a Hard Drive than the initial boot up?  Hardly anything.  In these tests, we used a very basic installation of Windows 98 Second Edition, which we installed on the same 10.2GB Quantum Hard Drive.  We booted up the computer freshly from a total shut down, and timed it's startup speed from pushing the power button, to the Network Login box.

 

 

The time difference of about 4 seconds between the two interfaces is hardly even noticeable when sitting in front of the computer.  Doc and Jeff wouldn't have even noticed a difference if it were for their stop watches.  Because of this, it's proven that in real world uses, where the hard drive wants lots of random different data, the seek time and rotational speed of the drive have more to do with the speed of the drive than anything else.  If our two testers were to use a 5400RPM drive in this testing, they are willing to bet with 10:1 odds that the difference would be even less noticeable, if there would be one at all.  The synthetic benchmarks can tell the difference between ATA/66 and ATA/33, but humans usually can't.  If you're buying a new system, it is recommended that you purchase at ATA/66 7200RPM drive along with a motherboard with ATA/66 included.  If you're thinking about upgrading your system to ATA/66, think again.  The cost involved with this is around $200.  If you really need to upgrade because you're doing graphics, try SCSI.  If you want to upgrade for the all-around speed, don't.  It's not worth the cost to get a few more seconds out of your boot time, and fractions of seconds on the load time of your favorite programs.  ATA/33 is good enough for today's slow mechanical drives, but as always, technology will advance, and maybe in 9 months or so, we will start to see drives that take true advantage of the bandwidth ATA/66 has to offer.

Ultra ATA/100

http://www.pcmech.com/show/harddrive/12/

 

As we look at our computers, we can see the speed of the processor increase steadily with time (though it seems to be going at a much accelerated pace in recent years). We also see the size of the hard drive, the power of the video card, and practically every other aspect of the computer get faster, stronger, bigger, and so on. With all of these components becoming greater in their own respects, it becomes possible to create programs which are more intensive on all areas of the computer. Take, for example, newer games. Many of them have installation routines which easily take up nearly a GB (sometimes more) on the hard drive. Much of this data is audio or video, and oftentimes the game, with its fancy graphics, will be rather processor intensive.

 

Now, having lots of multimedia on the drive is all well and good, and having it be processor intensive is the choice of the designer, but one thing needs to be accounted for. If you are trying to play a massive, high quality video, or perhaps sending lots of data to the processor, you need the capacity to transfer the data fast enough to not encounter a bottleneck.

 

This is where forward-looking companies like Quantum come in (perhaps Quantum should send me free stuff for that plug). They recently developed the Ultra ATA/100 interface for newer hard drives. The idea behind this is to increase the transfer rate between the hard drive and the computer itself. With this in mind, the goal is to eliminate data bottlenecks on even the largest of files, allowing for the transfer and storage of massive files, without having to deal with long waiting periods. And now for some technical aspects.

 

Quantum touts the main points of the specification to be data transfer rates of up to 100MB/s and full backward compatibility with ATA/33 and ATA/66. Another main point that Quantum has tried to make rather explicit is that the drives will be plug and play. This was somewhat of a problem for ATA/66 drives, as the controller often required drivers to fully utilize the abilities. With full compliance (already garnered) from companies like Intel, they managed to get full support in the chipsets of new and upcoming motherboards. Motherboards coming out towards the end of this year are already expected to be fully compliant with ATA/100, as are hard drives released in Fall of 2000.

 

Having a maximum transfer rate of 100MB/s seems excessive, but the goal with it is to be prepared for future drives, rather than to make as much use of it in present drives. The true goal is to always keep the transfer rate between the drive and the computer greater than the transfer rate within the drive itself. If the internal transfer rate is greater, then we have lag caused by lots of data ready to be transferred, some of it in the buffer, some of it still sitting in the middle of the drive, waiting to be taken to the buffer. With this being the case, drives have to be built which have larger buffers, causing them to be more expensive. So, in an effort to keep things flowing smoothly, if the drive to computer transfer rate is always greater than the internal data transfer rate, things will always be flowing to and from the computer, without the annoying bottlenecks that cause delays we all hate. As an example, today's fastest drives have internal transfer rates of around 56MB/s, just under the 66MB/s limit of ATA/66. Therefore, whenever the drive needs to send something to the computer, the path is not so crowded that nothing can get through.

 

The interface utilizes the same 40-pin, 80-wire IDE cable that was introduced with ATA/66. Every other wire leads nowhere to prevent line crosstalk.

 

Present recommendation is to stick with ATA/33, or ATA/66 if you have one of the fastest drives on the market. There is presently no need for anyone to have an ATA/100 controller, though the fact that the technology exists is good, as it will give way to newer, faster technology, keeping our computer oriented society moving forward.

RAID: Your Guide

http://www.pcmech.com/show/harddrive/296/

 

Two things blend together to make RAID more powerful than ever: An increasing number of die-hard, PC-loving speed- freaks and an ever-decreasing price of the hard drive. We're (for most of us) beyond the stage of thinking our hard drives are too small. We're beyond the stages of making due because a hard drive costs so much. But, for the PC enthusiast, we're not beyond the stage of saying, "Damn, that hard drive is too slow!".

 

There is where RAID comes in. Individually, most hard drives today are too slow. Regardless of how fast they are designed to be, with the speed of today's processor and other system components, hard drives today are a source of incredible bottleneck for a system. With RAID, we can blend the power of two or more hard drives together to accomplish great things.

 

What is it?

RAID stands for Redundant Array of Inexpensive Disks. This is actually a great name for it. And with the price decreasing like never before, the "Inexpensive" part of the name is now becoming a reality. Depending on the setup you choose for your RAID array, it can offer you increased performance by using the power of two hard drives as a single volume or simply using the redundancy of a second drive for increased data security. Just like designers do in mission-critical machines (building redundant systems in case of the failure of one), a RAID array can provide increased security in the event of the failure of one of the drives. I will get into the RAID types in a minute, but any good RAID array will use mirroring technology, meaning that whenever you write something to your primary drive, the RAID setup will simultaneously write the same info to the secondary disk, meaning you always have a duplicate copy. In the event one drive fails, you have an exact, working copy of your entire system on the second disk.

 

The word "array" usually implies a series of elements, each of a similar size and nature. Well, RAID is no different. The optmimum setup for a RAID array employs two identical hard drives. If one of your drives is a 7200 RPM drive, then its best to be sure the other one is also a 7200 RPM drive. The same goes for capacity. If you have one 20 gig drive and the other is a 10 gig drive, your 20 gig drive will only operate on the RAID array as a 10 gig drive. In the example preceding, that RAID array would operate at 5400 RPM if you had a 5,400 RPM drive paired up with the 7200 RPM drive. Summing up, your RAID array will always operate at the speed or capacity of the weakest or smallest drive. A chain is only as strong as its weakest link. So, obviously, if you're looking to set up a RAID setup, buy two identical drives.

 

As you might guess, you need a special controller to set up a RAID array. The controller handles the task of managing read/write requests to both drives, managing the mirroring, etc. On some operating systems, namely NT Server or Win2000, you can use the OS itself as a software-based controller. But, it is always better to install a separate, hardware-based PCI controller. The PCI controller handles all the work onboard, saving the CPU cycles that a software controller would use. Controller cards also come with software to allow you to monitor the status of the array.

 

Redundancy is the key to a RAID array, but regardless of whichever setup you employ, you will defintely use one or more of the following:

 

Striping
This is a RAID configuration that can offer huge performance gains. Data in a striped array is interleaved across all the drives in the array. Data is read and written on both drives at the same time. A good analogy would be this: Imagine having to write an essay on a sheet of paper. You can take a pen and write it. Now, imagine for a second that you were a mythological God or something and could write with both hands, nice and neat, at the SAME time. Imagine how fast you could write that paper now! This theory applies to a RAID array using striping. By splitting the data up and using both drives to read/write, it effectively doubles the speed.

 

The performance of a striped array is governed by the stripe width and stripe size. The width is equal to the number of drives in your array. To outline this, assume you need to write a 1 meg Word file to your RAID array. If you have two drives, then the stripe width is two. For purpose of clarifying, assume you will be writing this data in 50K chunks. That is 20 write cycles to write the entire Word file, 10 write cycles per drive. So, the first drive writes the first 50K, then the third, then the fifth, etc. At the same time, the other drive writes the second, then the fouth, etc. You can see that this setup would write the entire 1 meg file in about half the time of one drive. You can increase performance even more by adding another hard drive to the RAID array, thereby increasing the stripe width to 3.

 

The stripe size is basically the size of those chunks of data being written across the array. Default for an IDE configuration is usually 64K. Contrary to common sense, increasing the stripe size can have a negative impact on performance. See, if the data chunks are huge, then many times the parallel nature of RAID will not even be employed, because the chunks may be larger than the files themselves. This would lead to no better performance than a non-RAID setup. On the flip side, a stripe size that is too small will guarantee that your file will be broken up across the array (increasing performance) but increases the liklihood of small-time random accesses to the array, meaning your drives will likely be busier. As you can see, its a give-and-take thing.

 

Mirroring
With striping alone, you do not get any redundancy. The data is all split up amongst the drives in the array, so if you lose one of the drives, you're screwed. Mirroring is the other feature of RAID that comes to the rescue. The only problem is that with mirroring, you don't get striping. Mirroring is a simple concept: whatever you write to one drive, you write simultaneously to the other. Thus, you always have an exact duplicate of your data on the second drive. The cool parts of this come with the controller you decide to use. For example, most controllers will automatically sense a drive failure and instantly switch to the backup drive, meaning virtually no downtime. This is great for servers and other mission-critical machines. If the controller doesn't support this, it will most likely at least automatically transfer the data from the backup drive to the new drive.

 

Mirroring does give a small performance benefit as well. Since both drives contain similar data, the controller can read data from one drive while simultaneously requesting data from the copy. But, write speeds will slow down some, because the controller must write all data twice.

 

Parity
Parity is another type of redundancy built into some RAID arrays. Instead of simply making copies of everything, the RAID controller adds a parity bit to all binary info being written to the array. Basically, its just an extra bit of data appended onto the actual data. This series of parity bits is added up by the controller to equal either an even or an odd number. By analyzing this value, the controller can determine whether the information has been compromised in any way. If it has, it can replace the data automatically with data from the other drive.

 

Most parity setups use the XOR to do their magic. This is a type of Boolean logic, the eXclusive OR. Basically, it analyzes the series of 0's and 1's and returns either a TRUE or FALSE (even numbers are TRUE, odd is FALSE). By using this data, the controller can "fill in the blanks". Its like algebra. We know that 3 + 4 = 7. If you see an equation like 3 + __ = 7, you know the blank is supposed to be a 4. The XOR logic is used in this way to rebuild corrupted data on the array, thus maintaining integrity.

 

The more commonly used RAID levels are RAID 0, RAID 1, RAID 0+1, and RAID 5. Each "level" is simply a different configuration of the RAID standard, each providing certain benefits and performance parameters.

 

RAID 0

RAID 0 could be said to not be technically RAID. Why? Because it lacks the "R" - redundancy. RAID 0 is basically a RAID setup that employs the striping I talked about above. This setup requires at least two hard drives to be configured into a "striped set". RAID 0 is becoming increasingly popular amongst power users. As discussed before, this setup offers much higher read/write speeds than normal and will really help to speed up a computer. People who are into raw speed for gaming, multimedia, etc, will enjoy RAID 0. But, because it lacks the redundancy factor, it is not typically used in corporate, mission-critical environments. If one drive of the RAID 0 array dies, the whole array is screwed.

 

RAID 1

RAID 1 employs the mirroring capability discussed previously. It can, in some cases, provide a little performance benefit, but it is primarily used for redundancy, pure and simple. With RAID 1, you have the option of attaching a third drive to the controller. It acts as a spare drive. It is not part of the RAID array, but simply kicks in in the event that one of the drives fails. The controller would perform an automatic restore to the spare drive, notify you of the failure, and continue operating as though nothing happened. RAID 1 is used more on corporate networks andweb servers. Desktop users don't typically need it, although some who REALLY need that redundancy do use it on desktop machines.

 

RAID 0+1

RAID 0+1, as you might be able to tell from the name, gives you the best of both worlds. It can be costly, though, as it requires at least 4 hard drives to do it. Two of the drives are striped, as in a RAID 0 array, and the other two are mirrors of the first two. This is the only option for IDE users who want both the speed and the redundancy. Due to the cost of buying 4 hard drives plus a RAID controller, this is not the most popular option in town. It does, though, kick ass, and you will find desktop users and web server guys using this.

 

RAID 5

RAID 5 uses the high performance capability of striping with the increased integrity of the parity bit. The setup requires at least 3 drives. To see why it needs 3, see the discussion of parity above. By comparing the data on two of the drives, it can "fill in the blanks" on the third drive, just like solving an algabraic equation. This is what gives RAID 5 the security. Because both the data and parity info is spread out across all drives, it is often called "distributed parity".

 

RAID 5 is typically not an option for desktop users. It offers the best of all worlds, but typically only SCSI RAID controllers have the ability to handle it. This means IDE cannot be used, which in turn means this option will cost a crapload. RAID 5 is typically thought to be used in enterprise servers and the like.

 

JBOD
I love the name of this one - JBOD, "Just a Bunch of Drives". No kidding. This is barely RAID at all. It basically uses the controller to span two drives together into a single drive volume. When one of the disks fill up, it starts using the other one, transparently to the user. This setup will utilize all the space of the drives, which means you won't lose any space with differently sized drives placed on the array. On the flip side, though, it doesn't offer any redundancy or performance benefits. You will find that many controllers offer this as an option, although there's not a huge point in using it, in my opinion.

Simply put, IDE. Okay, well let me clarify my position.

 

Until more recently, an IDE RAID array would have been grounds for fun and laughter. IDE was slow and the whole setup just wasn't worth the hassle. But, we now have ATA/66 and, better yet, ATA/100 drives. And they are dirt cheap. Today, IDE RAID arrays are a great alternative to a SCSI array.

 

First, let me tell you, SCSI arrays are EXPENSIVE. A good SCSI RAID controller will set you back several hundred dollars. Add on top of that the need for two or more SCSI hard drives (depending on what RAID level you will be employing). Are you seeing the dollar signs yet? On the plus side, SCSI does offer a wider array of options and is faster. Also, in big server environments, IDE would be a bad option because the IDE design limits the number of drives to four. SCSI RAID can support up to 60.

 

IDE RAID is more affordable and quite fast. Many times, a good RAID 0 array using two decent speed IDE drives can outperform a high-end SCSI alone, while costing much less. For this reason, many users are finding IDE RAID 0 (or possibly RAID 0+1) a good way to go. Some of today's real powerful systems are now employing RAID arrays.

 

Setting up an IDE RAID array is not that difficult. There are some things you will need and some things you need to do first.

 

1. Make sure you have a valid, working system disk before doing anything. Create it and test it. Make sure it also has the necessary files to boot your CD-ROM. If you have a CD backup, you will need to get that CD-ROM working before you can proceed.
2. In # 1, I said "If you have a CD backup". In Step 2, I say, "Make a CD backup". Or some kind of backup. Unless you are setting up a simple RAID 1 array, your data on the first drive will be hosed. You are starting your system from scratch. So, before doing anything, make backups of everything you deem important. A more thorough solution would be to create an image file of your whole drive and back it up on a CD.
3. Wet, rinse, repeat step 2. (Just emphasizing.)
4. You'll need hard drives and a controller. You will need at least 2 hard drives, possibly more depending on the type of RAID you are doing. Get a RAID controller that matches your drives' specs.

 

Start playing:

1. Grab some standard IDE ribbon cables and connect your two hard drives to your RAID controller card. For two drive configurations, use one cable per drive, and attach each drive to its own channel on the card. For four-drive configurations, use a cable per 2 drives, just like you would with a standard master/slave setup.
2. Going with the previous step, you need to act as if each channel on the RAID controller is an IDE channel of your motherboard. Thus, if you only have two drives, each drive should be set to master. If you have four, each channel should have one master and one slave.
3. Install the controller into an open PCI slot.
4. Connect the hard drives to the case, as you would with any other drive installation. Be sure to connect the power leads, too.
5. After confirming everything is in place and ready to go, boot the PC. You will likely be brought to a configuration screen for your RAID controller. Here you will need to configure the array type, such as striped, mirrored, etc.
6. Insert your system disk into Drive A: and reboot.
7. Use FDISK to partition the array.
8. Likewise, format the array. All this is just like any normal installation.
9. Restore your programs and backup data.
10. Pat yourself on the back. Twice.

 

Things to Watch

There are two major points to keep in mind when installing RAID in your system.

 

First, your motherboard must have a good bus-mastering DMA sequencer on board. Bus-mastering is a technique which allows hardware to communicate to other hardware on the same bus without going through the processor. This reduces the load on the CPU. The DMA sequencer is what assigns your four of eight bus-mastered DMA channels to your PCI slots. Not only does your RAID controller have to be installed into one of these bus-mastered slots, but the DMA sequencer must be robust enough to handle it. This is kind of a trial and error thing, although some controller manufacturers will post on their web site a listing of tested motherboards that will work well.

 

Second, your RAID card is picky in that it wants to be in the very first bus-mastered slot on the motherboard. You will need your motherboard's manual to determine which slot is the first. Some boards count their bus numbers from the top down, others from the bottom up. Some start from the AGP slot and count down. So, since the RAID controller needs to be first, make sure it is. If you have a PCI video card, make sure the RAID controller is above it (or whichever slot is numbered first). If you're using AGP, pay attention to see if the first PCI slot below that shares an IRQ with the AGP video card. If it does, you'll need to move the RAID controller down a slot. The manual is your best reference. In a crunch, you can always use trial and error - remove all cards from system except for RAID and video, and move around until it works, then re-install all the other cards.

 

Conclusion

I hope you found this article useful. RAID is defintely a viable option for the speed freaks out there. Some of us even have a few hard drives lying around with decent specs. Popping a cheap RAID controller into a system and putting those drives to use could really improve your system's performance.

Common Hard Drive Problems

http://www.pcmech.com/show/harddrive/68/

 

If you’re like most people, you have either already ran out of space on your hard drive, or you are soon to do so. And you’ll probably go out and get a new hard drive, either new or used. The new ones usually come with software that set the drive up for you, by partitioning and formatting it. The used ones usually don’t. That’s where the trouble starts.

 

The most common problem I get from people trying to set up their hard drive is: "My (Larger than 2GB) Drive is only showing 2GB." The problem for that is usually in the Operating System (OS for short). The first version of Windows 95, for example, uses a file system called FAT16. That file system limits the size of the hard drive that is visible to the OS to only 2GB. So when you try to make that larger, it won’t let you. Plain and simple as that. You either must partition your hard drive into several 2GB partitions, or upgrade to an OS that with a file system that will support more than 2GB on a partition.

 

Another reason is because your BIOS has limits. 386 and 486 and lower end Pentium systems have limits of 512MB. Some Pentium Systems are limited to 2GB, and some of the newer ones, are limited to 8GB. It’s all in how the BIOS address the clusters on the Hard drive. It can be corrected with software, that comes with most new drives, like Western Digital’s EZ Drive, and Quantum’s Disk manager just to name a few. They take over where your real BIOS can’t perform, and then addresses the hard drive correctly

The next most common problem I get is "My hard drive says it’s 2GB, but Windows is saying it’s 1.86GB. Where’d that 90MB of space go?" Well, that problem is all in the numbers. The makers of the hard drive count 1MB as 1,000,000 Bytes. Windows counts 1MB as 1,048,576 bytes, a difference of 48,576 bytes. That adds up when you are talking 2,000MB. Let’s do the math.

Makers of hard drive says there are 2,000,000,000 bytes on the drive, so divide that by 1024 to get the number of kilobytes on the drive. Do that again to get the number of megabytes on the drive. Once more for the number of Gigabytes on the drive. You should get 1.862645149GB, or just 1.86GB, which is what Windows is thinking. That’s where your space went, in the numbers.

Another problem I am asked the answer for are a lot of FAT32 ones. "What is FAT32?" "Should I switch to FAT32?" "Can I switch to FAT32 and keep my data on the drive." "What OSs support FAT32."

Versions of Windows95 older than OSR2, as well as any DOS version, operate on a file system called FAT16 (or FAT12 in some cases). The existence of large hard drives has led to large partition sizes, which mean large cluster sizes and wasted space. Under FAT16, a smaller cluster size is better, because a small file takes up a whole cluster if there is even one byte in it; the leftover space is called "slack." FAT32 changed that.

FDISK in Windows 95 OSR2 or later will only allow you to put FAT32 on drives larger than 512MB. (Unless you use the /fprmt switch when starting FDISK) Inside FDISK, you must enable "large disk support," to choose FAT32. After exiting FDISK and rebooting, FORMAT the drive. NOTE that you must manually reboot after exiting FDISK, this is not automatic as in previous versions of FDISK. If you do not reboot between FDISKing and FORMATing, you will get strange-looking error messages.

As always, when you FDISK a drive, you will loose all data. But there are programs out there, like the one that comes with Windows 98, and Partition Magic, that will convert your drive to FAT32 without loosing your data.

With that, I hope that somehow, and someway, your Hard drive upgrades, and future problems, will be easily corrected.