Airflow UI Overview
The Airflow web interface is your control center. You use it to monitor workflows, trigger runs, view logs, and manage connections — all without touching the command line. This topic walks through every important section of the UI.
Logging In
Open your browser and go to http://localhost:8080. Enter the username and password you created during setup. After logging in, you land on the DAGs page — the main dashboard.
The DAGs Page (Main Dashboard)
The DAGs page lists all your workflows. Think of it as your project board. Each row is one workflow.
┌─────────────────────────────────────────────────────────────────┐ │ DAG Name │ Runs │ Schedule │ Last Run │ Status Summary │ ├─────────────────────────────────────────────────────────────────┤ │ daily_sales │ 42 │ 0 6 * * *│ 2h ago │ ●●●●●●● (7 ok) │ │ weekly_report │ 6 │ 0 8 * * 1│ 5d ago │ ●●●○●●● (1 fail)| │ etl_pipeline │ 10 │ @hourly │ 55m ago │ ●●●●●●●●●● ok │ └─────────────────────────────────────────────────────────────────┘
Each row shows how many times the workflow ran, its schedule, the last run time, and a visual summary of recent run results. Green circles mean success. Red circles mean failure.
Toggle Switch (Pause / Resume)
Each DAG row has a toggle switch on the left. Flip it off to pause a DAG so it stops running on schedule. Flip it on to resume. This is useful when you need to edit a workflow without deleting it.
Trigger DAG Button
Click the play button on a DAG row to run it immediately, regardless of its schedule. Use this to test a workflow on demand.
Inside a DAG: The Detail Views
Click any DAG name to open its detail page. You get several view tabs, each showing your workflow from a different angle.
Grid View
Grid View shows all past runs as columns and all tasks as rows. Color-coded squares tell you the status of every task in every run at a glance.
Tasks │ Run1 │ Run2 │ Run3 │ Run4 │ ───────────────┼──────┼──────┼──────┼──────┤ fetch_data │ ✅ │ ✅ │ ✅ │ ✅ │ clean_data │ ✅ │ ✅ │ ❌ │ ✅ │ save_to_db │ ✅ │ ✅ │ ⬜ │ ✅ │ send_email │ ✅ │ ✅ │ ⬜ │ ✅ │
In Run3, clean_data failed (❌). Because it failed, save_to_db and send_email never ran (⬜ = skipped or not started).
Graph View
Graph View draws your workflow as a flowchart. Boxes are tasks. Arrows show which task feeds into which.
[fetch_data]
|
v
[clean_data]
|
v
[save_to_db]
|
v
[send_email]
Click any box in Graph View to open that specific task's details, logs, and options like "Clear" or "Mark Success."
Calendar View
Calendar View shows a monthly calendar. Each day shows whether the DAG run for that day succeeded (green), failed (red), or did not run (grey). This helps you spot patterns — for example, if runs always fail on Mondays.
Gantt View
Gantt View shows how long each task took as horizontal bars on a timeline. This helps you find slow tasks that are making the entire workflow take longer than expected.
Task | 0s 5s 10s 15s 20s 25s 30s ─────────────┼───────────────────────────────────── fetch_data |[====] clean_data | [==========] save_to_db | [====] send_email | [==]
In this example, clean_data takes the longest. You know to optimize that task first if you want to speed up the pipeline.
Logs View
Click any task square in Grid View, then click "Log." Airflow shows you the full output of that task — every print statement, error message, and stack trace. Logs are your best debugging tool.
The Top Navigation Menu
DAGs
Returns you to the main dashboard at any time.
Browse
The Browse menu gives you access to historical data — all DAG runs, all task instances, and all log events. Use it to investigate what happened on a specific date or in a specific run.
Admin
The Admin menu holds the most important management pages:
- Connections: Store credentials for databases, APIs, and cloud services. You define them here once and reference them in your code by name.
- Variables: Store key-value pairs that your DAGs can read at runtime. Good for environment-specific settings.
- Pools: Limit how many tasks can run at the same time to avoid overloading a resource.
- XComs: See the small data values that tasks pass to each other during a run.
Security
The Security menu lets you manage users, roles, and permissions. You control who can view, edit, or trigger DAGs.
Task Instance States and Their Colors
| Color | State | Meaning |
|---|---|---|
| Green | Success | Task finished without errors |
| Red | Failed | Task encountered an error |
| Yellow | Running | Task is currently executing |
| Light Blue | Queued | Task is waiting for a worker slot |
| Orange | Up for Retry | Task failed and will retry soon |
| Pink | Skipped | Task was intentionally skipped |
| White / Grey | No Status | Task has not been scheduled yet |
How to Clear and Re-run a Failed Task
When a task fails, you can fix the problem and re-run just that task without re-running the entire DAG:
- Go to the DAG's Grid View
- Click the red square of the failed task
- Click Clear
- Confirm. Airflow resets that task and re-runs it from where it failed
This saves time — you do not repeat successful tasks, only the failed one.