Dagster Cloud

1.13.0 · active · verified Fri Apr 10

Dagster Cloud is a fully-managed cloud platform for Dagster, providing hosted orchestration, development tools, and operations dashboards. The `dagster-cloud` Python library offers a client to interact with the Dagster Cloud API programmatically, manage deployments, and configure agents. It is currently at version 1.13.0 and follows the Dagster core library's release cadence, typically releasing new versions monthly.

Warnings

gotcha Authentication to Dagster Cloud requires setting `DAGSTER_CLOUD_ORGANIZATION_ID` and `DAGSTER_CLOUD_API_TOKEN` environment variables. Without these, client initialization will fail or lead to unauthorized access errors.
Fix: Set `DAGSTER_CLOUD_ORGANIZATION_ID` and `DAGSTER_CLOUD_API_TOKEN` in your environment. API tokens can be generated in the Dagster Cloud UI.
breaking The `dagster-cloud` library must be compatible with the core `dagster` library. Significant version mismatches (e.g., pre-1.0 `dagster` with 1.x `dagster-cloud`) can lead to import errors, runtime exceptions, or unexpected behavior due to API changes and data model evolution.
Fix: Always install `dagster` and `dagster-cloud` within the same major version range. Refer to the official Dagster documentation for version compatibility guidelines.
gotcha Changes to deployment models (e.g., agent-based, hybrid, serverless) often require updates to `dagster-cloud` configuration or even CLI commands. Relying on outdated deployment patterns with newer `dagster-cloud` versions can lead to deployment failures or inefficiencies.
Fix: Regularly review the Dagster Cloud deployment documentation and ensure your deployment configuration and `dagster-cloud` CLI usage align with the recommended patterns for your chosen deployment model.

Install

pip install dagster-cloud Install dagster-cloud

Imports

DagsterCloudClient

from dagster_cloud.client import DagsterCloudClient

DagsterCloudCodeLocation
```
from dagster_cloud.dagster_cloud_definitions import DagsterCloudCodeLocation
```
The definition location changed with version 1.x to be more explicit about cloud-specific definitions.

Quickstart

This quickstart demonstrates how to initialize the `DagsterCloudClient` and use it to interact with your Dagster Cloud organization. It retrieves and prints the names of all available workspaces. Ensure `DAGSTER_CLOUD_ORGANIZATION_ID` and `DAGSTER_CLOUD_API_TOKEN` environment variables are set for authentication.

import os
from dagster_cloud.client import DagsterCloudClient

# Ensure these environment variables are set:
# DAGSTER_CLOUD_ORGANIZATION_ID
# DAGSTER_CLOUD_API_TOKEN

organization_id = os.environ.get('DAGSTER_CLOUD_ORGANIZATION_ID', 'your_org_id_here')
api_token = os.environ.get('DAGSTER_CLOUD_API_TOKEN', 'your_api_token_here')

if not organization_id or not api_token:
    print("Error: DAGSTER_CLOUD_ORGANIZATION_ID and DAGSTER_CLOUD_API_TOKEN must be set.")
    exit(1)

try:
    client = DagsterCloudClient(
        organization_id=organization_id,
        api_token=api_token
    )
    
    # Example: List available workspaces
    workspaces = client.get_workspaces()
    print(f"Found {len(workspaces)} workspaces in Dagster Cloud:")
    for ws in workspaces:
        print(f"- {ws.name} (ID: {ws.id})")
        
except Exception as e:
    print(f"An error occurred: {e}")

view raw JSON →