IBM Z Workload Scheduler
IBM's batch scheduler for z/OS, formerly Tivoli Workload Scheduler for z/OS and, before that, OPC.
Also written TWS, OPC, Tivoli Workload Scheduler for z/OS, IZWS, current plan
IBM Z Workload Scheduler is IBM's scheduling product for mainframe batch. It has had several names, and people still use all of them: OPC, its original name, and Tivoli Workload Scheduler for z/OS, usually shortened to TWS.
Work is defined in applications, groups of operations, where each operation is usually a job. Each application has run cycles that say which days it runs, and dependencies that say which operations must finish first, including operations in other applications.
From these definitions, the scheduler builds plans. The long-term plan looks weeks ahead. The current plan covers the next day or so and is what actually drives the work: operations become ready when their time arrives and their predecessors finish, and the scheduler submits them. Staff watch and control the current plan through ISPF screens or a web interface, and can hold, rerun or complete operations by hand.
As with any scheduler, its definitions are the real description of how the batch fits together. When you need to know what runs before or after a job, look in the scheduler, not just in the JCL.
Where can I learn this hands-on?
Right here, at ZedBank. You take a junior job at a simulated bank and learn each part of the work by doing it, with Zed beside you. There's nothing to install.
Start at ZedBankRelated
- Batch schedulerSoftware that submits production jobs automatically in the right order, at the right time, based on what has already succeeded.
- Job dependencyA rule that a job must wait for something else first, usually another job finishing successfully.
- Control-MBMC's batch scheduling product, which decides when each job runs based on times, dependencies and conditions.
- Batch windowThe period, usually overnight, in which batch work has to finish before the business needs the systems back.
- Critical pathThe chain of dependent batch jobs that decides when the whole batch run finishes; any delay on it delays the finish.
- RerunRunning a job again after it failed or produced wrong results, either from the beginning or from a chosen step.