IBM i Knowledge Base

IBM i Operating System Explained

Beyond the definition, IBM i works differently under the hood than Windows, Linux, or UNIX: work management runs through subsystems and jobs, changes are applied through PTFs, and system startup follows an IPL rather than a simple reboot. This is how it actually operates day to day.

What is IBM i? covers the definition and architecture at a high level. This article goes one layer deeper into how the operating system actually runs day to day: how work is organized, how the system starts up, and how IBM delivers and applies fixes.

Subsystems and Work Management

IBM i organizes running work into subsystems, which are OS-managed environments that control how batch jobs, interactive sessions, and communications work are scheduled and resourced. A typical IBM i system runs several subsystems concurrently, for example one handling interactive 5250 sessions, one handling batch processing, and one handling TCP/IP server jobs. Administrators tune subsystem storage pools and job queues to prioritize critical workloads, which is a core part of IBM i performance management.

Libraries and the Object Model

IBM i groups related objects, programs, files, and other resources, into libraries, which function as both an organizational structure and a search mechanism. When a user or job runs, IBM i resolves object references through a library list, an ordered search path the system checks to find the correct version of a program or file. This differs fundamentally from a directory-tree filesystem: every object on IBM i has a type, and the OS enforces what operations are valid against that type at the system level.

The IPL Process

Starting or restarting an IBM i system is called an IPL (Initial Program Load), the IBM i equivalent of a boot sequence. IPLs can be normal (a controlled shutdown and restart) or abnormal (following an unplanned outage), and administrators distinguish between attended and unattended IPLs depending on whether manual intervention is required during startup. Because IBM i systems often run for extended periods without a full IPL, planning IPL windows is a routine part of IBM i operations, typically scheduled around maintenance activity.

PTF Maintenance

IBM delivers IBM i fixes and updates as PTFs (Program Temporary Fixes), packaged individually or in cumulative groups. IBM i shops apply PTFs on a maintenance schedule, testing groups of fixes in a non-production environment before applying them to production, since some PTFs require an IPL to take effect. Staying current on PTFs is a standard element of IBM i support and security hygiene, distinct from a full OS version upgrade such as moving from IBM i 7.4 to 7.5.

Frequently Asked Questions

What is a subsystem on IBM i?
A subsystem is an OS-managed environment that controls how a category of work, such as batch jobs, interactive sessions, or communications jobs, is scheduled and resourced. IBM i systems typically run several subsystems at once.
What is an IPL on IBM i?
IPL stands for Initial Program Load, the IBM i term for starting or restarting the system. It is functionally similar to a reboot on other platforms but follows IBM i's specific startup sequence and can be normal, abnormal, attended, or unattended.
What is a PTF?
A PTF (Program Temporary Fix) is how IBM delivers fixes and updates to IBM i. PTFs are typically applied in tested groups on a maintenance schedule, and some require an IPL to take effect.
How does IBM i organize files and programs?
IBM i groups objects into libraries and resolves references through a library list, an ordered search path. Every object has a system-enforced type, which differs from the directory-tree filesystem model used on Windows or Linux.