Beam has not been initialized. Call beam.Init() before…

Error message

Beam has not been initialized. Call beam.Init() before pipeline construction.

What it means

The Dataflow runner requires beam.Init() to have been called so flags (project, region, staging location, etc.) are parsed and the environment is configured. Execute panics immediately if beam.Initialized() is false, because building and submitting a job without that configuration cannot proceed safely.

Solutions

Call beam.Init() (which parses flags) at the start of main before building the pipeline
Ensure flag.Parse() runs — beam.Init uses the standard flag package
In tests, call beam.Init() in TestMain or skip Dataflow-dependent tests
Run the pipeline via beam.Run(ctx, runner, p) after initialization rather than invoking dataflow.Execute directly

Example fix

// before
func main() {
    p := beam.NewPipeline()
    dataflow.Execute(ctx, p)
}
// after
func main() {
    beam.Init()
    p := beam.NewPipeline()
    beam.Run(ctx, "dataflow", p)
}

Defensive patterns

Strategy: validation

Validate before calling

if !beam.Initialized() {
    beam.Init()
}

Try / catch

func runDataflow(ctx context.Context, p *beam.Pipeline) (pr beam.PipelineResult, err error) {
    defer func() {
        if r := recover(); r != nil {
            err = fmt.Errorf("dataflow execute panicked: %v", r)
        }
    }()
    if !beam.Initialized() { beam.Init() }
    return dataflow.Execute(ctx, p)
}

Prevention

Always call beam.Init() as the first statement in main
Never invoke runner Execute functions directly; use beam.Run
In tests, call beam.Init() from TestMain
Wrap pipeline construction so it only happens after flag parsing

When it happens

Trigger: Constructing a beam.NewPipeline() and running it with the Dataflow runner without ever calling beam.Init() (typically from the main function before flag.Parse).

Common situations: New Go Beam users wiring pipelines in tests or libraries where beam.Init() lives in main() that never runs, or calling dataflow.Execute directly instead of through beam.Run after flag parsing.

Related errors

AI-assisted analysis of apache/beam@12126d8942 (2026-09-13). Data as JSON: /api/errors/54a082e8f58358f8. Report an issue: GitHub.

Appendix: source

Thrown at sdks/go/pkg/beam/runners/dataflow/dataflow.go:197

}

var unique int32

// Helper function finding first non empty string. Used for handling alias options.
func firstNonEmpty(values ...*string) *string {
	for _, value := range values {
		if *value != "" {
			return value
		}
	}
	return values[0]
}

// Execute runs the given pipeline on Google Cloud Dataflow. It uses the
// default application credentials to submit the job.
func Execute(ctx context.Context, p *beam.Pipeline) (beam.PipelineResult, error) {
	if !beam.Initialized() {
		panic("Beam has not been initialized. Call beam.Init() before pipeline construction.")
	}

	edges, nodes, err := p.Build()
	if err != nil {
		return nil, err
	}
	streaming := !graph.Bounded(nodes)

	beam.PipelineOptions.LoadOptionsFromFlags(flagFilter)
	opts, err := getJobOptions(ctx, streaming)
	if err != nil {
		return nil, err
	}

	// (1) Build and submit
	// NOTE(herohde) 10/8/2018: the last segment of the names must be "worker".
	id := fmt.Sprintf("go-%v-%v", atomic.AddInt32(&unique, 1), time.Now().UnixNano())

View on GitHub (pinned to 12126d8942)