Command Line Interface

Overview

The Dagu CLI provides all the necessary commands to manage and execute DAGs (workflows) directly from the terminal. It allows you to start, stop, retry, and monitor workflows, as well as manage the underlying scheduler and web server.

Complete Reference

For the full CLI command reference, see CLI Commands Reference.

Basic Usage

dagu [global options] command [command options] [arguments...]

Getting Help

# General help
dagu --help

# Command-specific help
dagu start --help

# Show version
dagu version

Essential Commands

Running Workflows

Start a Workflow

# Basic execution
dagu start my-workflow.yaml

# Interactive DAG selection (when no file is specified)
dagu start

# With named parameters (use -- separator)
dagu start etl.yaml -- DATE=2024-01-01 ENV=prod

# With positional parameters
dagu start my-workflow.yaml -- value1 value2 value3

# Queue for later
dagu enqueue my-workflow.yaml

Stop a Running Workflow

# Stop currently running workflow
dagu stop my-workflow

# Stop specific run
dagu stop --run-id=20240101_120000 my-workflow

# Can also use file path
dagu stop my-workflow.yaml

Restart a Workflow

# Restart latest run
dagu restart my-workflow

# Restart specific run
dagu restart --run-id=20240101_120000 my-workflow

Retry Failed Workflow

# Retry specific run (run-id is required)
dagu retry --run-id=20240101_120000 my-workflow

# Can also use file path
dagu retry --run-id=20240101_120000 my-workflow.yaml

Monitoring Workflows

Check Status

# Check latest run status
dagu status my-workflow

# Check specific run status
dagu status --run-id=20240101_120000 my-workflow

# Can also use file path
dagu status my-workflow.yaml

View Status of a DAG run

# Check detailed status and output
dagu status my-workflow.yaml

# Note: For detailed logs, use the web UI at http://localhost:8080
# or check log files in the configured log directory

Server Commands

Start Everything

# Start scheduler, web UI, and coordinator service (default: localhost:8080)
dagu start-all

# Custom host and port
dagu start-all --host=0.0.0.0 --port=9000

# Custom DAGs directory
dagu start-all --dags=/path/to/directory

Start Scheduler Only

# Run just the scheduler (no UI)
dagu scheduler

# Custom DAGs directory
dagu scheduler --dags=/opt/workflows

Start Web UI Only

# Run just the web server (no scheduler)
dagu server

# Custom host and port
dagu server --host=0.0.0.0 --port=9000

# Custom DAGs directory
dagu server --dags=/path/to/directory

Distributed Execution Commands

Start Coordinator

# Start the coordinator gRPC server
dagu coordinator

# Custom host and port
dagu coordinator --coordinator.host=0.0.0.0 --coordinator.port=50055

# With TLS
dagu coordinator \
  --peer.cert-file=server.pem \
  --peer.key-file=server-key.pem

The coordinator service manages task distribution to workers for distributed execution with automatic service discovery and health monitoring.

Start Worker

# Start a worker that polls for tasks
dagu worker

# With labels for capability matching
dagu worker --worker.labels gpu=true,memory=64G,region=us-east-1

# With custom worker ID and concurrency
dagu worker \
  --worker.id=gpu-worker-01 \
  --worker.max-active-runs=50

Workers automatically register in the service discovery system and poll the coordinator for matching tasks based on their labels.

Interactive DAG Selection

When you run dagu start without specifying a DAG file, an interactive selector appears:

dagu start

Features:

• Browse available DAGs with filtering
• Enter parameters interactively
• Confirm before execution

Advanced Usage

Queue Management

# Add to queue
dagu enqueue my-workflow.yaml

# Add to queue with custom ID
dagu enqueue --run-id=custom-001 my-workflow.yaml

# Add to queue with parameters
dagu enqueue my-workflow.yaml -- KEY=value

# Remove from queue (requires DAG-name:run-id format)
dagu dequeue --dag-run=my-workflow:custom-001

Working with Parameters

Parameters can be passed in multiple ways:

# Positional parameters (use -- separator)
dagu start my-workflow.yaml -- param1 param2 param3

# Named parameters (use -- separator)
dagu start my-workflow.yaml -- KEY1=value1 KEY2=value2

# Mixed (use -- separator)
dagu start my-workflow.yaml -- param1 KEY=value param2

CLI Configuration

Global Options

Option	Description	Default
--config	Config file path	~/.config/dagu/config.yaml
--log-level	Log verbosity	info
--log-format	Output format	text
--quiet	Suppress output	false

See Also

• Explore the REST API for programmatic access
• Set up the Web UI for visual monitoring
• Learn workflow syntax to build complex DAGs
• Configure distributed execution for scaling workflows