pids

Process identifiers, commonly abbreviated as PIDs, are numeric labels assigned by an operating system to individual processes to identify and manage them during execution. A PID is typically valid only for the lifetime of its process and is unique within the system (or within a given namespace). PIDs are used by system utilities and programming APIs to refer to processes, and many commands display the PID of each listed process.

In Unix-like systems, the kernel assigns a PID when a new process is created. The current process can learn its own identifier with system calls such as getpid(), while the process’s parent can be identified with getppid(). Information about a process can be examined under the /proc filesystem, where each process has a directory named by its PID. The first user-space process usually has PID 1, and the kernel traditionally reserves PID 0.

PIDs are finite and recycled; once a process terminates, its PID may later be reassigned to a new process. If a terminating process has not yet been waited on by its parent, it may become a zombie, leaving its PID occupied until the parent collects the exit status. In Linux and other systems that support containers, PID namespaces allow separate sets of PIDs to exist in different isolation domains, enabling processes inside containers to have PIDs that do not collide with those in the host.

Other uses of the acronym PID exist outside operating systems; for example, PID also stands for proportional–integral–derivative controller, a common feedback-control algorithm in engineering. In Windows environments, processes also have PIDs that can be queried and manipulated by appropriate APIs.