import dashboard

[iceangel86868686-bot]

Hi, I ran into a problem when importing dashboards, the charts are not displayed until you click save to edit dataset.

And you don't need to edit anything, just click the "save" button
How to do it massively when there are a lot of datasets

[FrancescoCastaldi]

This issue occurs because when dashboards are imported from a ZIP/YAML export, the dataset metadata (physical database column types, calculated metrics, and temporal flags) is imported from the export schema, but the underlying database connection in the target environment has not yet synchronized column types with the target database engine.

When you click "Edit Dataset" and then "Save", Superset triggers UpdateDatasetCommand / sync_columns(), which re-queries the database metadata via SQLAlchemy and populates table_columns and sql_metrics.

How to synchronize all datasets in bulk

You do not need to edit them manually one by one. You can automate this using either the Superset CLI or a short Python administrative script.

Method 1: Administrative CLI Command

Execute the refresh-datasources CLI command inside your Superset container or virtual environment:

# In Docker Compose
docker compose exec superset superset refresh-datasources

# Or targeting a specific database connection
docker compose exec superset superset refresh-datasources --database-name "YourDatabase"

Method 2: Python Script (Bulk Refresh via Superset App Context)

Run this script inside the Superset container using superset shell or python:

from superset.app import create_app
app = create_app()

with app.app_context():
    from superset import db
    from superset.connectors.sqla.models import SqlaTable

    datasets = db.session.query(SqlaTable).all()
    print(f"Syncing metadata for {len(datasets)} datasets...")

    for ds in datasets:
        try:
            ds.fetch_metadata()
            print(f"Successfully refreshed: {ds.table_name} (ID: {ds.id})")
        except Exception as e:
            print(f"Error refreshing {ds.table_name}: {e}")

    db.session.commit()
    print("Bulk metadata sync complete.")

Method 3: REST API Refresh Loop

If you prefer using the REST API from CI/CD, call the refresh endpoint for each dataset ID:

# Obtain your JWT token via /api/v1/security/login, then:
curl -X POST "http://localhost:8088/api/v1/dataset/<DATASET_ID>/refresh" \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json"

Running either Method 1 or 2 will synchronize all columns and metrics across all imported datasets at once.

[FrancescoCastaldi]

A quick precision regarding the REST API endpoint:
In Superset 6.x, the dataset refresh route (superset/datasets/api.py) is exposed strictly via the HTTP PUT method:
curl -X PUT "http://localhost:8088/api/v1/dataset/<DATASET_ID>/refresh"
Using POST returns HTTP 405 Method Not Allowed. Additionally, when importing dashboards or datasets programmatically, passing sync_columns=true in the import form data ensures columns and metrics are synchronized automatically upon import.