The Basics of Servers: Their Components and Use

The Basics of Servers: Their Components and Use
Page content

A server can refer to either software of hardware that has been designed to perform special functions thereby increasing the flexibility or capability of information processing and the consequential use of that information. In the case of hardware, a server is a computer that has been set aside from regular desktop usage to run server-specific applications.

Servers differ from desktop computers in fundamental ways because they are typically run applications under heavy workloads. These computers are expected to be turned on and work for long periods of unattended time, often twenty-four hours a day and seven days a week. Some servers are turned on and aren’t scheduled for maintenance for an entire year. For example, a web server is a computer that stores website information and must be ready to handle requests for a website at all times. When a request for a web page is made by an Internet user, the web server calculates the request and returns the requested information to the client, in this case a collection of pictures, text, scripts, etc. that are necessary to display the webpage on the client’s computer.

Hardware

Hardware requirements for servers differ depending on their application. To a greater extent than home computers, there is no one best configuration. The CPU speed of a server is usually not as important a concern as throughput. Throughput is defined as the amount of data that can be processed by a computer. Although the CPU is a big contributor to throughput, other aspects such as RAM speed and size, BUS speed, and hard drive read/write speeds contribute significantly to throughput. Often, servers have multiple CPUs to increase throughput rather than as a means to increase raw processing power.

Servers often have higher RAM requirements than desktop computers to satisfy the throughput paradigm. The graphical ability of a server is rarely a useful feature since servers are not required to display graphical output to a monitor. Even in a gaming server, the key is information throughput, not graphical ability. Servers sometimes lack a Graphical User Interface (GUI) or even run headless; meaning without a monitor attached. Reduced or no grahpical requirements free up processing resources, allowing the server to be more efficient in performing its intended tasks.

Servers usually have higher networking requirements to serve hundreds of users simultaneously where fast network connections are matched to the needs of the server. Often, this includes much faster upload bandwidth than is found in homes and offices.

Space can be a consideration when many servers occupy a small location. The form factors of these computers vary greatly from the typical desktops found in homes and offices. For example, rack-mounted servers allow many servers to reside in a vertical structure from floor to ceiling. These systems range in height from an inch to about a foot. These servers are small because they often lack their own optical drives, power supplies, and storage. Instead, several rack-mounted servers share these peripherals. Some larger servers have redundant power supplies so the failure of one power supply does not result in the server’s shutdown.

Storage in servers is often a system of drives rather than single drives. A RAID (Redundant Array of Independent Disks) allows for various configurations. For example, one type of RAID stripes data over several drives so that the computer can use multiple drives as if they were one. Another type allows for redundancy in case one drive fails, while another speeds up seek and read times so data flows more quickly. Still more types offer some combination of these features

A server room is a special room or “closet” that is typically artificially cooled to reduce overheating caused by a multitude of servers within close proximity to one another. These rooms also reduce noise levels created when so many servers are running at the same time. An Uninterruptible Power Supply (UPS) is used to supply power in case of a power outage. These UPSs are capable of running the servers for hours or even days without outside power.

Conclusion

Although servers often share some components with home and office computers, the purpose of a server is to serve information to many simultaneous users. Form follows function with some servers lacking key components found in home or office computers. Lack of monitors, graphical ability, graphical user interfaces, and peripherals allow servers to occupy much less space thereby making setup and maintenance easier and performance more efficient.