Hard Drive 101 - Basics Guide
Every computer needs memory, and most modern desktops and laptops rely on hard disk drive (HDD) technology to store files. In fact, if you are a regular computer user, there is a good chance that you have operated hundreds of hard drives over the past few decades, and while other technologies are gaining popularity, hard drives are by far the most commonly used option for both personal and commercial data storage.
Hard drives can be thought of as a series of precise mechanical components that work together to store digital information with magnetic charges. These charges are stored on disks, referred to as platters, and each platter has a set of heads that flies across its surface to read and write data.
Of course, hard drives are not simple devices, and this definition leaves out quite a bit of important information.
This article will explain some of the basic attributes of a typical hard drive, including access speeds, storage capacity, interfaces and common storage configurations that use multiple hard drives.
How Hard Drive Capacity is Determined and Reported
Hard drives gained popularity because they offer relatively high maximum storage capacities at a low price as compared to NAND flash and other storage technologies.
Many modern hard drives have capacities of a terabyte or more. This means that you have a terabyte of free space from the moment you plug the drive into your computer - but not in the way you might think.
Digital data is stored in bits, and as storage capacities have expanded, manufacturers have used larger and larger units to describe hard drive storage. It is very important to note that since the beginning of the consumer hard drive industry, manufacturers have defined a kilobyte as 1,000 bytes (called the metric prefix) while software developers have defined a kilobyte as 1,024 bytes (the binary prefix). This is an important designation, because it directly affects the usable capacity of a hard drive. If you purchase a 1 terabyte hard drive, it will only have about 909 gigabytes of usable data when viewed from your computer. A 2 terabyte hard drive will only have 1.82 terabytes of usable data. The larger the storage device, the bigger the disparity between the reported capacity and the usable capacity.
Keep this in mind when shopping for data storage devices or evaluating hard drive errors. When you purchase a drive, make sure that the device will have enough space for your software, documents, videos and other files while still leaving about 30 percent of the space free to ensure optimal operational efficiency.
Hard Drive Operating Information and Important Numbers to Consider
Hard drive manufacturers also provide other operating numbers to demonstrate the performance of their products. Many of these numbers can be useful when diagnosing hard drive performance issues.
Examples of performance statistics include:
Below are some of the examples of hard drive performance based on statistical information.
- Rotation Speed (RPM) - This is the mechanical speed of the hard drive, expressed as the number of rotations per minute. Most desktop hard drives operate around 7200 RPMs, although speeds of up to 15,000 RPMs are available. Laptop hard drives are often slower (but not necessarily); other common speeds are 4200 and 5400 RPMs. Faster drives usually have faster access speeds, which can result in a smoother user experience. However, the mechanical speed of a hard drive is not the only factor that determines its overall efficiency.
- Power Consumption - The amount of power that the hard drive uses during typical operation. Drives with lower rotation speeds usually consume less power than hard drives with fast rotation speeds.
- Seek Time - The time that a hard drive’s read/write heads take to locate a specific track on the platters (usually expressed as an average). Seek time can be thought of as the drive’s mechanical efficiency - if seek time slows down, it could indicate a problem with the electronics, actuator arm, head assembly or another hard drive component.
- Data Transfer Rate - The actual rate at which a hard drive communicates data to the motherboard of the computer. This is an excellent characteristic to look for when comparing similar hard drives’ real performance, as the faster data transfer rate will usually offer a better computing experience (but not necessarily better dependability or durability). Some manufacturers will show a “maximum data transfer rate,” which uses an ideal set of operating conditions.
- Operating Noise - Like most mechanical devices, hard drives can make sounds when operating normally. Many manufacturers share the operating noise of their hard drives, expressed as dBA (decibels adjusted). Quieter drives may be preferable for certain applications.
Hard Drive Interface Types
Hard drives communicate with computers through a printed circuit board (PCB) outfitted with an appropriate interface, which is sometimes referred to as a bus type. Each interface type has different characteristics, especially in relation to data transfer rates.
The most common hard drive interface types include:
SATA
Serial Advanced Technology Attachment or SATA is by far the most common interface for newer hard drives. New revisions of the technology can handle transfer speeds of up to 1.97 gigabytes per second. SATA also has relatively small transfer and power cables as compared to PATA.
ATA / PATA
Also known as EIDE, IDE and by several other acronyms, Parallel Advanced Technology Attachment (PATA) was commonly used through the early 2000s. Unlike SATA, this interface used large ribbon cables for data transfers, which could create space constraints in some desktop computer towers. PATA’s maximum data transfer speed was 133 megabytes per second. While PATA drives are still manufactured, SATA is a superior technology and a better option for most computers.
SAS
SAS stands for “Serial Attached Small Computer System Interface,” an update to older SCSI technologies. SAS uses SATA-like connections, and many computers with SAS capabilities can also read and write data from SATA drives. SAS offers transfer speeds of up to 12 gigabytes per second. These transfer speeds, coupled with SAS’s ability to receive standard SCSI commands, make the interface an excellent option for servers and other high-demand applications.
Fibre Channel
Introduced after SCSI, Fibre Channel (FC) is usually used for storage area networks (SAN). It offers fast transfer speeds, but contrary to the name, it does not use fibre optics to transfer data at the disk level. Older drives may use other interface types, including parallel SCSI (often referred to simply as SCSI), which was commonly used for server and enterprise applications through the late 1990s and early 2000s. External hard drives typically use one of the interfaces listed above, but add an external USB or Firewire port that allows them to connect to computers easily. These ports include USB, eSATA, Thunderbolt and IEEE 1394 (FireWire).
Choosing from Different Types of Hard Drives
There are four basic categories of hard drives: desktop, laptop, server and external drives.
Desktop Hard Drives
Desktop hard drives are 3.5-inches in length, while laptop hard drives measure at 2.5-inches (they are also much slimmer on average).
Laptop Hard Drives
While laptop hard drives can be used in desktop computers with special mounting hardware, they are relatively expensive as compared to 3.5-inch drives, so there are few reasons to mount a 2.5-inch drive in a desktop computer except in special circumstances.
External Hard Drives
External hard drives consist of a standard 2.5-inch or 3.5-inch hard drive in a portable enclosure with its own built-in electronics.
Server Hard Drives
Server hard drives usually have the same 3.5-inch form factor, but have additional features (such as SAS connections) that make them ideal for constant operation in demanding applications. They are typically more expensive than standard desktop hard drives.
Solid-State Drives and Hybrid Hard Drives
In recent years, solid-state drives (SSD) have become a popular choice for home computer users. SSD technology offers several performance advantages over traditional hard drives, including faster read/write speeds and resistance to damage from physical shocks and vibrations.
Solid-state drives use electronic circuits to store data. Unlike standard hard drives, they do not store magnetic charges, and they do not have mechanical components. However, they cannot offer perfect protection against data loss - the NAND flash memory used by solid-state drives can wear down over time, which is why most newer drives employ wear-leveling technologies to extend operating life. Currently, SSD technology is about as reliable as HDD technology, but SSDs that are used constantly can pose a higher threat of data loss. The other major disadvantage of SSD is that it is expensive as compared to traditional hard drives, and storage capacities are much lower. Many computer users choose to use SSD for their operating systems and software and HDD for long-term data storage, which can provide the advantages of both technologies.
Some manufacturers also offer hybrid drives (sometimes referred to as SSDHD), which have fairly large NAND flash storage areas and mechanical hard drive components. These drives only spin their platters to access data when absolutely necessary, which should theoretically extend the operating life of the drive while reducing energy consumption and improving data access speeds.
RAID and Other Configurations with Multiple Hard Drives
Multiple hard drives can be configured for use as a single storage device. This usually requires a special piece of hardware called a controller, but software can act as a controller for smaller multi-drive configurations. Some external hard drives with large capacities actually consist of several hard drives linked together with a hardware controller.
Most computers that use multiple hard drives use a configuration called RAID (Redundant Array of Inexpensive/Independent Disks). There are several RAID levels, which offer different benefits in terms of performance and redundancy (the term “redundancy” refers to the array’s ability to withstand hard drive failures without losing data).
Some of the most popular types of RAID are listed below.
RAID 1
Also known as “mirroring,” this configuration consists of multiple hard drives that are identical to one another. Data is written to all of the drives in the array at the same time. In a two-drive RAID 1, if either drive fails, the other drive has the same set of data. RAID 1 offers the same data access speeds as a standard hard drive.
RAID 0
RAID 0 is not technically a RAID, since it does not provide any redundancy. Data is striped across the member disks, so each drive has an equal amount of the information. This allows for fast access speeds, since all of the drives can write at the same time. However, because there is no redundancy, a single hard drive failure will result in data loss.
RAID 5
This is the most common RAID configuration used for web servers and larger storage systems. It distributes a parity across the member drives, which means that all of the data exists in at least two locations; a RAID 5 can sustain a hard drive failure without losing any data. A RAID 5 requires at least three hard drives. It offers performance improvements over a RAID 1, but usually requires a controller card for parity calculations.
RAID 6
Similar to RAID 5, but with another parity block. This allows for more redundancy. RAID 6 is often used in larger systems where downtime is unacceptable. The added parity block may mean reduced performance, but RAID 6 can be just as fast as RAID 5 depending on its implementation.
Nested RAID
This term refers to RAID configurations that combine two implementation technologies for improved performance and redundancy. For example, a RAID 10 consists of a RAID 1 and a RAID 0, creating a stripe of mirrored drives. This allows for fast write speeds with some fault tolerance. Other common nested RAID levels include RAID 50 (combines RAID 5 and RAID 0) and RAID 01 (the opposite of a RAID 10, a mirror of striped drives).
JBOD
Stands for “Just a Bunch of Disks,” a non-RAID architecture that allows users to write data across multiple drives or to any of the drives independently. Businesses rarely use JBOD, since it does not offer performance improvements or redundancy, but it can offer a simple way to store large amounts of data in some applications.
Some manufacturers have their own proprietary RAID configurations, which may offer improved redundancy or protection against data corruption. In any case, it is still important to follow proper backup procedures when using a RAID array; no RAID configuration can protect perfectly against data loss.
Why Do Hard Drives Fail?
Every hard drive fails eventually, usually due to the normal stress that accompanies day-to-day operation. Hard drive mechanical components operate at fast speeds with incredible precision, and over time, this can wear out motors and other physical components. A drive might also fail due to electronic damage, physical shocks or for other environmental reasons.
Many hard drives use Self-Monitoring, Analysis and Reporting Technology (abbreviated as SMART) to detect the early onset of performance issues. SMART can basically show when certain parameters are out of line, predicting a drive failure before it occurs. However, it is not perfect, and while it is a useful tool for diagnosing impending read/write head failures and other issues, it is not perfectly reliable.
This is why regular data backup is so important: since every drive fails and failures are somewhat unpredictable, backup is the best way to protect your most important files. If you do not backup your hard drive and you experience a failure, data recovery is usually an option, but in order to improve your chances of a positive result, keep your system powered down and contact a professional provider as soon as possible.