parallel

⚡ Parallel Execution Example (Worker Pool)

This example demonstrates how to optimize engine performance by evaluating multiple rules simultaneously using the Worker Pool system.

Concepts Illustrated

1. Parallel vs Serial

By default, the engine evaluates rules one after the other. If some rules use slow dynamic facts (e.g., API calls, DB queries), the total execution time is the sum of all these times.

With parallel execution:

• Rules are distributed to a group of workers.
• Multiple slow facts can be calculated at the same time.
• Determinism: Although the evaluation is parallel, the events (Success/Failure) are triggered sequentially according to the priority order to ensure predictable behavior.

2. Configuration (WithParallelExecution)

You can enable this feature when creating the Engine:

engine := gre.NewEngine(
    gre.WithParallelExecution(5), // 5 concurrent workers
)

Performance comparison in this example

The example simulates 5 rules depending on a fact that takes 100ms to calculate:

• Serial: ~500ms (100ms * 5)
• Parallel (5 workers): ~100ms (all rules run simultaneously)

Execution

go run main.go