Snapshots

Trigger pipelines from the dashboard, SDK, CLI, or REST API.

A Pipeline Snapshot is an immutable snapshot of your pipeline that includes the pipeline DAG, code, configuration, and container images. Snapshots enable you to trigger pipeline runs without direct access to the codebase—from the ZenML Pro dashboard, Python SDK, CLI, or REST API.

Running snapshots is a ZenML Pro feature. For comprehensive documentation including advanced usage patterns, see the main Snapshots documentation.

Why Use Snapshots?

Snapshots solve common production challenges:

• Data Scientists can experiment with different parameters without modifying code
• MLOps Engineers can schedule retraining or integrate with CI/CD systems
• Stakeholders can trigger model training through the dashboard
• External Systems can invoke pipelines via REST API calls

Requirements

Snapshots require a remote stack with at least:

• Remote orchestrator
• Remote artifact store
• Container registry

Local stacks cannot run snapshots.

Platform Engineers: For snapshots to work in Hybrid or Self-hosted deployments, you must configure the workload manager on your Workspace Server. See Enable Snapshot (Workload Manager) Support for the Workspace Server for the required environment variables and Kubernetes RBAC setup.

Creating Snapshots

From the CLI

zenml pipeline snapshot create <PIPELINE-SOURCE-PATH> --name=<SNAPSHOT-NAME>

You can also specify a configuration file and stack:

zenml pipeline snapshot create <PIPELINE-SOURCE-PATH> \
    --name=<SNAPSHOT-NAME> \
    --config=<PATH-TO-CONFIG-YAML> \
    --stack=<STACK-ID-OR-NAME>

From Python SDK

from zenml import pipeline

@pipeline
def my_pipeline():
    ...

snapshot = my_pipeline.create_snapshot(name="production-training")

From the Dashboard

1. Navigate to a pipeline run
2. Click ... in the top right corner
3. Select + New Snapshot
4. Enter a name and click Create

Running Snapshots

From the CLI

zenml pipeline snapshot run <SNAPSHOT-NAME-OR-ID>

# With custom configuration
zenml pipeline snapshot run <SNAPSHOT-NAME-OR-ID> --config=config.yaml

From Python SDK

from zenml.client import Client

# Get the snapshot
snapshot = Client().get_snapshot("<NAME-OR-ID>")

# Optionally modify configuration
config = snapshot.config_template
config.steps["my_step"].parameters["my_param"] = new_value

# Trigger the run
Client().trigger_pipeline(
    snapshot_name_or_id=snapshot.id,
    run_configuration=config,
)

From the Dashboard

1. Click Run a Pipeline on the Pipelines page, or navigate to a snapshot and click Run Snapshot
2. Modify configuration using the built-in editor or upload a YAML file
3. Click Run

From REST API

curl -X 'POST' \
  '<YOUR-ZENML-SERVER>/api/v1/pipeline_snapshots/<SNAPSHOT-ID>/runs' \
  -H 'Authorization: Bearer <TOKEN>' \
  -H 'Content-Type: application/json' \
  -d '{
    "run_configuration": {
      "steps": {
        "train_model": {
          "parameters": {"model_type": "rf"}
        }
      }
    }
  }'

For REST API authentication, use Personal Access Tokens or Service Accounts.

Deleting Snapshots

zenml pipeline snapshot delete <SNAPSHOT-NAME-OR-ID>

Or via Python:

from zenml.client import Client
Client().delete_snapshot(name_id_or_prefix="<SNAPSHOT-NAME-OR-ID>")

Important Notes

After upgrading your ZenML server, you need to recreate your snapshots. Snapshots are tied to specific server versions and may not work correctly after an upgrade.

Related Documentation

• Snapshots - Full Documentation - Complete reference with advanced usage
• Trigger Pipelines from External Systems - Tutorial for CI/CD integration
• Workspace Server Configuration - Configure the workload manager that powers snapshots
• Service Accounts - Set up API authentication for automated triggers