A program is an executable file that is held in storage. Storage refers to devices or media that can retain data for relatively long periods of time (e.g., years or even decades), such as hard disk drives (HDDs), optical disks and magnetic tape. This contrasts with memory, whose contents can be accessed (i.e., read and written to) at extremely high speeds but which are retained only temporarily (i.e., while in use or only as long as the power supply remains on).
An executable file is a binary file (i.e., a file at least part of which is not plain text) that has been compiled (i.e., converted using a special type of program called a compiler) from source code into machine machine code, which is a pattern of bytes that can be read directly by a central processing unit (CPU). Source code is the version of software as it is originally written (i.e., typed into a computer) by a human in plain text (i.e., human readable alphanumeric characters). A CPU is the main logic unit of a computer.
A program is a passive entity until it is launched, and a process can be thought of as a program in action. Processes are dynamic entities in that they are constantly changing as their machine code instructions are executed by the CPU. Each process consists of (1) system resources that are allocated to it, (2) a section of memory, (3) security attributes (such as its owner and its set of permissions) and (4) the processor state.
The processor state includes the contents of its registers and physical memory addresses. Registers are a very small amount of very fast memory that is built into a processor in order to speed up its operations by providing quick access to commonly used values. A memory address is a location in memory.
An alternative definition of a process is the execution context of a running program, i.e., all of the activity in the current time slot in the CPU. A time slot, also called a time slice or a quantum, is the length of time that each process is permitted to run in the CPU until it is preempted (i.e., replaced) by another process in a time sharing operating system.
Linux and other Unix-like operating systems have been designed from the ground up as complete time sharing systems, that is, as both multitasking and multi-user systems. A multitasking system is one that allows multiple processes to operate seemingly simultaneously without interfering with each other, and a multi-user system allows multiple users to use the system simultaneously, with each having the illusion of being the sole user.
This intricate but robust time sharing capability is made possible by the ability of the system to both retain many processes in memory at the same time and switch between them fast enough to make it appear as though they are all running simultaneously. If one process crashes (i.e., stops functioning), it will usually not cause other processes to crash because each process runs in its own protected memory space (i.e., area of memory) and is not capable of interacting with other processes except through secure mechanisms managed by the kernel (i.e., the core of the operating system).
Programs and processes are distinct entities. Thus, in a multitasking operating system, multiple instances of a single program can be executing simultaneously, and each instance is a separate process (or processes). For example, if seven users, each with their own keyboard and display device, decide to run the vi text editor at the same time, there will be seven separate instances of vi, each a separate process, although they will all share the same executable file. A single user can likewise simultaneously run seven instances of vi, or some other program.
Another, compatible, definition of a process, for those familiar with the C programming language (in which the kernels and numerous other programs in Unix-like operating systems are written), is the collection of data structures that completely describe how far the execution of the program has progressed. A data structure is a way of storing data in a computer so that it can be used efficiently.
Fortunately, it is not necessary for ordinary users to fully comprehend these definitions in order to understand the basic concept of processes and to know how to use them to control their login sessions (i.e., use of a computer after entering the correct username and password) and make their work more efficient.
The Process Life Cycle
This results in the creation of init, which is the first process of the session and which becomes the ancestor of all other processes created during that session. The role of init is to read the entries in the file /etc/inittab and execute various programs according to that file. This includes starting the getty process on each of the login terminals, which eventually provides the designated shell for each user.
A shell is a program that provides the traditional, text-only user interface for Unix-like operating systems. Its primary function is to read commands that are typed into a console (i.e., an all-text display mode) or terminal window (an all-text window in a GUI) and then execute (i.e., run) them. A command is an instruction telling a computer to do something, such as start a program.
A program can be started automatically or by a user typing in the name (and correct path if necessary) of the program at the command line (i.e., all-text display mode) and then pressing the ENTER key. This causes the program to be read into memory and executed by the kernel. Some programs create a single process when launched, such as ls (which is used to show the contents of a directory), whereas others, such as OpenOffice (an increasingly popular and open source office suite), initiate a series of processes.
In Unix-like operating systems, each process is given a unique number, referred to as a process identification (PID), when it is created, and this number is used by the system to reference the process. Each process is guaranteed to have a unique PID, which is always a non-negative integer. init always has a PID of 1 because it is always the first process on the system. A very large PID does not necessarily mean that there are anywhere near that many processes on a system, because such numbers are often a result of the fact that PIDs are not immediately reused in order to prevent possible errors.
While a process is running, it can spawn (i.e., give birth to) other processes. Spawning is accomplished through the use of a system call termed a fork (because it splits in two). System calls are clearly defined, direct entry points into the kernel through which processes request services from the kernel.
The first step in spawning a new process is for an existing process to create an identical copy of itself. This copy is then transformed into the new process, and it, in turn, can create additional processes, thereby resulting in multiple generations of processes (i.e., parents spawn children which spawn grandchildren). Analogies can be made with the filesystem hierarchy of Unix-like systems and also with the object hierarchy in an object-oriented programming language (such as Java, in which all classes are descendants of the class named Object).
As is virtually everything else running in a Unix-like operating system, the shell is also a process. (The big exception is the kernel, which is a set of routines that resides continuously in memory and to which all processes have access.) When a user types in a command, the shell spawns a process that executes that command. Unless the user specifies otherwise, the shell typically waits for this child process to be completed before it displays the prompt again to indicate that it is ready for a new command. A prompt, also referred to as a command prompt, is a short text message at the start of each line on a console or terminal window.
If a process is suspended (i.e., temporarily not in use), it becomes eligible for swapping (i.e., transferring) to the swap partition in order to free up space in the main memory for other processes.
During its lifetime, a process will utilize a variety of system resources. They include (1) the processor to run its instructions, (2) the memory to hold it and its data, (3) files within the filesystem and (4) physical devices on the system. The operating system must keep track of each process and the resources it uses in order to manage it and the other processes efficiently, i.e., so that no one process monopolizes the processor or memory.
The ps command is used to list the currently running processes and their PIDs. At a bare minimum, two processes will be shown, the shell (usually bash on Linux) and ps, which itself is a process and which dies as soon as its output is displayed. Usually, there will be many more. The following will provide a full listing of the current processes:
The -a option tells ps to list the processes of all users on the system rather than just those of the current user. The -u option tells ps to provide detailed information about each process. The -x option adds to the list processes that have no controlling terminal, such as daemon that are started during booting. In contrast to most commands, the hyphen preceding the option(s) with ps is optional.
As the number of processes can be quite long and occupy more than a single screen, the output of ps aux can be piped (i.e., transferred) to the less command, which lets it be viewed one screenful at a time. The output can be advanced one screen forward by pressing the SPACE bar and moved one screen backward by pressing the b key.
Among the information that ps aux provides about each process is the user of the process, the PID, the percentage of CPU used by the process, the percentage of memory used by the process, VSZ (virtual size in kilobytes), RSS (real memory size or resident set size in 1024 byte units), STAT (the process state code), the starting time of the process, the length of time the process has been active and the command that initiated the process. The process state codes include D, uninterruptable sleep; N, low priority; R, runnable (on run queue); S, sleeping; T, traced or stopped; and Z, defunct (zombie).
The processes can also be viewed with the pstree command, which can be used as follows to list all of the processes currently on the system in the form of a tree diagram:
The addition of the -p option will also show the PIDs:
The processes that are directly connected to the main stem (i.e., a vertical line extending downward from init along the left hand edge of the screen) of the tree are listed by default in alphabetic order. This is in contrast to ps, which by default lists the processes in the order in which they were created. It can be seen that pstree itself is also listed as a process.
There are a number of reasons that a user would want to control processes, possibly the most common of which is to close a program that has frozen or crashed. This can be accomplished by using the lethal-sounding kill command.
For example, if the Mozilla web browser freezes and it cannot be closed by using ordinary keyboard commands or mouse clicks, it can in many cases be closed by first using
Another reason that a user might want to control processes is to make use of job control, a feature of the shell that facilitates the handling of multiple processes. Job control can be used to switch processes between the foreground and the background, and it allows programs to be started initially in the background.
Running a job in the background is typically done when its execution is expected to take a long time and in order to free the issuing terminal after entering the command. (Some processes are not suitable for running in the background, such as text editors, which occupy the full console or terminal window screen.) Starting a program in the background is accomplished by typing its name followed by an ampersand. For example, typing
at the command line starts gftp, an open source FTP (file transfer protocol) program that can be used to send files between computers, in the background. This frees the console or terminal window for use by other commands while gftp is downloading or uploading programs.
A process that is running in the foreground can be suspended by simultaneously pressing the CTRL and z keys and can be terminated by simultaneously pressing the CTRL and c keys. The command bg reactivates a suspended program in the background, and the command fg puts a suspended program or a program that is running in the background into the foreground.
Daemons are a class of processes that run continuously in the background, rather than under the direct control of a user. The term is derived from the ancient Greek word daimon, which refers to a supernatural being that is intermediate between a human and a god, or similar to a guiding spirit. Daemons are generally easy to recognize because their names end with the letter d.
Daemons are usually launched automatically while a computer is booting up and then wait in the background until their services are required. They typically respond to hardware activity, to network requests or to other programs by performing specified tasks. They can also configure hardware (such as the daemon devfsd, which can provide intelligent management of device entries in the device filesystem on some Linux systems), run scheduled tasks (e.g., crond) and perform a variety of other functions.
Another example is the secure networking daemon, xinetd (eXtended InterNET services Daemon), which is usually launched during booting and listens passively until a program, such as FTP or telnet, requests a connection.
On the Microsoft Windows operating systems, functions comparable to those of daemons are provided by processes called services. However, the term daemon is now sometimes used with regard to those operating systems as well.
Created May 21, 2004. Last updated June 12, 2006.