Summary of Introduction to Operating Systems.

Summary

Some years ago an operating system was defined as the software that controls the hardware, but the landscape of computer systems has evolved significantly since then, requiring a more complicated description. To increase hardware utilization, applications are designed to execute concurrently. However, if these applications are not care- fully programmed, they might interfere with one another. As a result, a layer of software called an operating system separates applications (the software layer) from the hard- ware they access.

When a user requests that the computer perform an action (e.g., execute an application or print a document), the operating system manages the software and hardware to produce the desired result. Operating systems are primarily resource managers—they manage hardware, including processors, memory, input/output devices and communication devices. The operating system must also manage applications and other software abstractions that, unlike hardware, are not physical objects.

Operating systems have evolved over the last 60 years through several distinct phases or generations that correspond roughly to the decades. In the 1940s, the earliest electronic digital computers did not include operating systems. The systems of the 1950s generally executed only one job at a time, but used techniques that smoothed the transition between jobs to obtain maximum utilization of the computer system. A job constituted the set of instructions that a program would execute. These early computers were called single-stream batch-processing systems, because programs and data were submitted in groups or batches by loading them consecutively onto tape or disk.

The systems of the 1960s were also batch-processing systems, but they used the computer’s resources more efficiently by running several jobs at once. The systems of the 1960s improved resource utilization by allowing one job to use the processor while other jobs used peripheral devices. With these observations in mind, operating systems designers developed multiprogramming systems that managed a number of jobs at once, that number being indicated by the system’s degree of multiprogramming.

In 1964, IBM announced its System/360 family of computers. The various 360 computer models were designed to be hardware compatible, to use the OS/360 operating system and to offer greater computer power as the user moved upward in the series. More advanced operating systems were developed to service multiple interactive users at once. Timesharing systems were developed to support large numbers of simultaneous interactive users.

Real-time systems attempt to supply a response within a certain bounded time period. The resources of a real-time system are often heavily under-utilized. It is more important for real-time systems to respond quickly when needed than to use their resources efficiently.

Turnaround time—the time between submission of a job and the return of its results—was reduced to minutes or even seconds. The value of timesharing systems in support of program development was demonstrated when MIT used the CTSS system to develop its own successor, Multics. TSS, Multics and CP/CMS all incorporated virtual memory, which enables programs to address more memory locations than are actually provided in main memory, which is also called real memory or physical memory.

The systems of the 1970s were primarily multimode timesharing systems that supported batch processing, time- sharing and real-time applications. Personal computing was in its incipient stages, fostered by early and continuing developments in microprocessor technology. Communications between computer systems throughout the United States increased as the Department of Defense’s TCP/IP communications standards became widely used—especially in military and university computing environments. Security problems increased as growing volumes of information passed over vulnerable communications lines.

The 1980s was the decade of the personal computer and the workstation. Rather than data being brought to a central, large-scale computer installation for processing, computing was distributed to the sites at which it was needed. Personal computers proved to be relatively easy to learn and use, partially because of graphical user interfaces (GUI), which used graphical symbols such as windows, icons and menus to facilitate user interaction with pro- grams. As technology costs declined, transferring information between computers in computer networks became more economical and practical. The client/server distributed computing model became widespread. Clients are user computers that request various services; servers are computers that perform the requested services.

The software engineering field continued to evolve,a major thrust by the United States government being aimed especially at providing tighter control of Department of Defense software projects. Some goals of the initiative included realizing code reusability and a greater degree of abstraction in programming languages. Another software engineering development was the implementation of processes containing multiple threads of instructions that could execute independently.

In the late 1960s, ARPA, the Advanced Research Projects Agency of the Department of Defense, rolled out the blueprints for networking the main computer systems of about a dozen ARPA-funded universities and research institutions. ARPA proceeded to implement what was dubbed the ARPAnet—the grandparent of today’s Inter- net. ARPAnet’s chief benefit proved to be its capability for quick and easy communication via what came to be known as electronic mail (e-mail). This is true even on today’s Internet, with e-mail, instant messaging and file transfer facilitating communications among hundreds of millions of people worldwide.

The ARPAnet was designed to operate without centralized control. The protocols (i.e., set of rules) for communicating over the ARPAnet became known as the Transmission Control Protocol/Internet Protocol (TCP/IP). TCP/IP was used to manage communication between applications. The protocols ensured that messages were routed properly from sender to receiver and arrived intact. Eventually, the government decided to allow access to the Internet for commercial purposes.

The World Wide Web allows computer users to locate and view multimedia-based documents (i.e., documents with text, graphics, animations, audios or videos) on almost any subject. Even though the Internet was developed more than three decades ago, the introduction of the World Wide Web (WWW) was a relatively recent event. In 1989, Tim Berners-Lee of CERN (the European Center for Nuclear Research) began to develop a technology for sharing information via hyperlinked text documents. To implement this new technology, He created the Hyper Text Markup Language (HTML). Berners-Lee also implemented the Hyper- text Transfer Protocol (HTTP) to form the communications backbone of his new hypertext information system, which he called the World Wide Web.

Hardware performance continued to improve exponentially in the 1990s. Inexpensive processing power and storage allowed users to execute large, complex programs on personal computers and enabled small to mid-size companies to use these economical machines for the extensive database and processing jobs that earlier had been dele- gated to mainframe systems. In the 1990s, the shift toward distributed computing (i.e., using multiple independent computers to perform a common task) rapidly accelerated. As demand for Internet connections grew, operating sys- tem support for networking tasks became standard. Users at home and in large corporations increased productivity by accessing the resources on networks of computers.

Microsoft Corporation became dominant in the 1990s. Its Windows operating systems, which borrowed from many concepts popularized by early Macintosh operating systems (such as icons, menus and windows), enabled users to navigate multiple concurrent applications with ease.

Object technology became popular in many areas of computing. Many applications were written in object-oriented programming languages, such as C++ or Java. In object-oriented operating systems (OOOS), objects repre- sent components of the operating system. Object-oriented concepts such as inheritance and interfaces were exploited to create modular operating systems that were easier to maintain and extend than operating systems built with pre- vious techniques.

Most commercial software is sold as object code. The source code is not included, enabling vendors to hide proprietary information and programming techniques. Free and open-source software became increasingly common in the 1990s. Open-source software is distributed with the source code, allowing individuals to examine and modify the software before compiling and executing it. The Linux operating system and the Apache Web server are both free and open source.

In the 1980s, Richard Stallman, a developer at MIT, launched the GNU project to recreate and extend most of the tools for AT&T’s UNIX operating system. Stallman created the GNU project because he disagreed with the concept of paying for permission to use software. The Open Source Initiative (OSI) was founded to further the benefits of open-source programming. Open-source software facili- tates enhancements to software products by permitting anyone in the developer community to test, debug and enhance applications. This increases the chance that subtle bugs, which could otherwise be security risks or logic errors, will be caught and fixed. Also, individuals and cor- porations can modify the source to create custom software that meets the needs of a particular environment.

In the 1990s, operating systems became increasingly user friendly. The GUI features that Apple had built into its Macintosh operating system in the 1980s were widely used in many operating systems and became more sophisticated. “Plug-and-play” capabilities were built into operating systems, enabling users to add and remove hardware components dynamically without manually reconfiguring the operating system.

Middleware is software that links two separate applications, often over a network and often between incompatible machines. It is particularly important for Web services because it simplifies communication across multiple architectures. Web services encompass a set of related standards that can enable any two computer applications to communicate and exchange data via the Internet. They are ready- to-use pieces of software on the Internet.

When the IBM PC appeared, it immediately spawned a huge software industry in which independent software vendors (ISVs) were able to market software packages for the IBM PC to run under the MS-DOS operating system. If an operating system presents an environment conducive to developing applications quickly and easily, the operating system and the hardware are more likely to be successful in the marketplace. Once an application base (i.e., the combination of the hardware and the operating system environment in which applications are developed) is widely established, it becomes extremely difficult to ask users and software developers to convert to a completely new applications development environment provided by a dramatically different operating system.

Operating systems intended for high-end environments must be designed to support large main memories, special-purpose hardware, and large numbers of processes. Embedded systems are characterized by a small set of specialized resources that provide functionality to devices such as cell phones and PDAs. In these environments, efficient resource management is the key to building a successful operating system.

Real-time systems require that tasks be performed within a particular (often short) time frame. For example, the autopilot feature of an aircraft must constantly adjust speed, altitude and direction. Such actions cannot wait indefinitely—and sometimes cannot wait at all—for other nonessential tasks to complete.

Some operating systems must manage hardware that may or may not physically exist in the machine. A virtual machine (VM) is a software abstraction of a computer that often executes as a user application on top of the native operating system. A virtual machine operating system man- ages the resources provided by the virtual machine. One application of virtual machines is to allow multiple instances of an operating system to execute concurrently. Another use for virtual machines is emulation—the ability to use software or hardware that mimics the functionality of hardware or software not present in the system. By pro- viding the illusion that applications are running on different hardware or operating systems, virtual machines promote portability—the ability for software to run on multiple platforms—and many other benefits.

A user interacts with the operating system via one or more user applications. Often, the user interacts with an operating system through a special application called a shell. The software that contains the core components of the operating system is referred to as the kernel. Typical operating system components include the processor sched- uler, memory manager, I/O manager, interprocess commu- nication (IPC) manager, and file system manager.

Almost all modern operating systems support a multiprogrammed environment in which multiple applications can execute concurrently. The kernel manages the execution of processes. Program components, which execute independently but use a single memory space to share data, are called threads.

When a process wishes to access an I/O device, it must issue a system call to the operating system. That system call is subsequently handled by a device driver—a software component that interacts directly with hardware—often containing device-specific commands and other instructions to perform the requested input and output operations.

Users have come to expect certain characteristics of operating systems, such as efficiency, robustness, scalability, extensibility, portability, security and protection, interactivity and usability.

In a monolithic operating system, every component is contained in the kernel. As a result, any component can directly communicate with any other. Monolithic operating systems tend to be highly efficient. A disadvantage of monolithic designs is that it is difficult to determine the source of subtle errors.

The layered approach to operating systems attempts to address this issue by grouping components that perform similar functions into layers. Each layer communicates exclusively with the layers immediately above and below it.

In a layered approach, a user process’s request may need to pass through many layers before completion. Because additional methods must be invoked to pass data and control from one layer to the next, system throughput decreases compared to that with a monolithic kernel, which may require only a single call to service a similar request.

A microkernel operating system architecture pro- vides only a small number of services in an attempt to keep the kernel small and scalable. Microkernels exhibit a high degree of modularity, making them extensible, portable

and scalable. However, such modularity comes at the cost of an increased level of intermodule communication, which can degrade system performance.

A network operating system runs on one computer and allows its processes to access resources such as files and processors on a remote computer.A distributed operating system is a single operating system that manages resources on more than one computer system. The goals of a distributed operating system include transparent performance, scalability, fault tolerance and consistency.