Autostart - Automatic startup

Enhanced Definition

In the z/OS environment, "autostart" refers to the capability of the operating system or associated automation software to automatically initiate the execution of specific tasks, started tasks (STCs), or address spaces during the Initial Program Load (IPL) process or under predefined conditions, without requiring manual intervention. Its primary purpose is to ensure the immediate availability of critical system components and applications following a system restart or recovery event.

Key Characteristics

- IPL Integration: Autostart mechanisms are typically configured within system initialization parameters (e.g., IEASYSxx member in SYS1.PARMLIB) or through advanced system automation products.
- Started Tasks (STCs): Most commonly, autostart applies to initiating critical started tasks such as JES2/JES3, VTAM, TCP/IP, DB2, CICS, IMS, and other essential system utilities or applications.
- Conditional Startup: Advanced automation tools can implement conditional autostart, where a task only starts if specific resources are available, or other prerequisite tasks have successfully completed their initialization.
- Order of Execution: The sequence in which tasks are autostarted is crucial to satisfy inter-task dependencies (e.g., TCP/IP must be active before applications that rely on its services). This order is managed via IEASYSxx parameters or automation rules.
- Recovery and Resilience: Autostart is a fundamental component of system recovery strategies, minimizing downtime by automatically bringing essential services back online after an outage.
- Configuration: Configuration involves specifying program names or procedure names, often within PROCLIB members, which are then referenced by IEASYSxx parameters or automation rules.

Use Cases

- System IPL: Automatically starting core z/OS components like JES2/JES3, VTAM, TCP/IP, and SMF during every system IPL to establish basic system functionality and network connectivity.
- Database and Transaction Managers: Initiating DB2 subsystems, IMS control regions, and CICS regions upon system startup to make enterprise applications immediately available to users.
- Middleware Services: Automatically launching MQSeries queue managers, WebSphere Application Server for z/OS controllers, and other critical middleware components required for application integration.
- Automation Agents: Starting automation product agents (e.g., IBM System Automation for z/OS, TWS for z/OS) that are responsible for monitoring, managing, and reacting to other system resources and applications.
- Security Services: Ensuring security managers like RACF, ACF2, or Top Secret are active early in the IPL process to enforce access controls and system integrity.

Related Concepts

Autostart is intrinsically linked to the IPL (Initial Program Load) process, as it dictates which services become active immediately after a system restart. It heavily relies on Started Tasks (STCs) and their corresponding JCL procedures defined in PROCLIB, which provide the execution environment for these services. System Automation products (like IBM System Automation for z/OS) significantly extend basic z/OS autostart capabilities by providing sophisticated dependency management, health checking, and automated recovery actions beyond simple startup. It is also crucial for the high availability of Subsystems such as DB2, CICS, and IMS.

Best Practices:

Prioritize Critical Services: Configure essential system services (JES, VTAM, TCP/IP, security) to autostart first, followed by critical applications based on their established dependencies.
Document Autostart Procedures: Maintain comprehensive documentation of all autostarted tasks, their interdependencies, and the specific configuration members involved (e.g., IEASYSxx, automation rules).
Utilize Automation Products: For complex and dynamic environments, leverage robust system automation products to manage autostart sequences, dependencies, health checks, and automatic restarts for enhanced resilience and reduced manual intervention.
Monitor Startup Logs: Regularly review system logs (SYSLOG, JOBLOGs of STCs) during and after IPLs to promptly identify any failed autostarts or startup-related issues.
Test Autostart Configurations: Periodically test autostart configurations in a controlled test environment to validate the correct startup order and functionality after simulated system restarts or outages.
Parameterize for Flexibility: Use symbolic parameters within PROCLIB