Workload Manager (WLM)

The workload manager is a component of z/OS® that provides the ability to manage multiple workloads at the same time within one z/OS image or across multiple images. When using WLM, you do not need to do any tuning or issue any commands. The kernel uses WLM to create child processes while running in goal mode.

When programs issue fork() or spawn(), the BPXAS PROC found in SYS1.PROCLIB is used to provide a new address space. For a fork(), the system copies one process, called the parent process, into a new process, called the child process. The forked address space is provided by WLM. Figure 1 shows how a fork() creates a new process.
Figure 1. How fork() creates a new process
The fork process copies one process into a new process, which is called the child process.

Existing MVS™ address space types such as TSO, STC, Batch, and APPC can request z/OS UNIX services. When one of those address spaces makes its first request to the z/OS kernel, the kernel dubs the task; that is, it identifies the task as a z/OS UNIX process. There are two types of processes: user processes, which are associated with a user, and daemon processes, which perform continuous or periodic system-wide functions such as a web server.

Daemons are programs that are typically started when the operating system is initialized and remain active to perform standard services. Some programs are considered daemons that initialize processes for users even though these daemons are not long-running processes. Examples of daemons are as follows:
  • cron, which starts applications at specific times
  • inetd, which provides service management for a network
  • rlogind, which starts a user shell session when one is requested, using a remote rlogin command
In similar systems, initialization typically starts a telnet daemon to perform terminal services.Daemons are not restarted if they stop. You can restart them in any of several ways:
  • The z/OS operator can restart daemons using a cataloged procedure. For more information, see Starting daemons.
  • A system programmer can restart the daemon from a shell.
  • You can use automation products such as Tivoli® NetView® for z/OS to notice daemons terminating and then restart them using cataloged procedures.

A process can have one or more threads; a thread is a single flow of control within a process. Application programmers create multiple threads to structure an application in independent sections that can run in parallel for more efficient use of system resources.

For more information about threads, see pthread_create (BPX1PTC, BPX4PTC) in z/OS UNIX System Services Programming: Assembler Callable Services Reference.