Skip to main content

Global Variables

Last updated on

Global variables are inputs that are defined at the process (release) level and shared across all phases and activities.

What are Global Variables?​

Global variables are:

  • Process-Wide: Available in all phases and activities
  • Highest Scope: Broadest variable scope
  • Release-Specific: Unique to each release
  • Configurable: Can be set per release

Examples​

Following are examples of global variables:

  • Release version or release name
  • Account/org/project identifiers (where needed)
  • Change request ID applying to the full release
  • Feature flag rollout strategy used across phases

Characteristics​

  • Highest-level scope: Available everywhere in the release
  • Typically stable: Usually stable across the lifecycle of that release execution
  • Evaluated via expression engine: Evaluated via the shared pipeline expression engine

Benefits of Variable Scopes​

Together, the three variable scopes provide:

  • Activity variables: Fine-grained, per-step configuration
  • Phase variables: Per-phase context, reducing duplication
  • Global variables: Release-wide context, enabling consistent naming, evidence, and correlations

These are stored and resolved by Release Orchestration and the orchestration engine to create the final executable graphs.

Defaults and overrides​

Global variables can have default values and can be overridden for a specific execution when needed.

Best Practices​

Meaningful Names​

Use descriptive names:

  • Recommended: deployment_timeout
  • Avoid: timeout1

Documentation​

Document variables:

  • Purpose
  • Expected values
  • Default values
  • Usage examples

Default Values​

Provide sensible defaults:

  • Common use cases
  • Safe values
  • Well-documented
  • Overridable