🔄 Execution Lifecycle in CCP

An Execution is the process of running a Method with user-provided inputs.
It captures the full context, metadata, environment, and outputs of that run.

📦 What Is an Execution?

An Execution is triggered when a user fills the form of a Method and runs it.
This generates a full execution bundle that includes:

Inputs and parameters
Method descriptor
Infrastructure and runtime details
Output files
Metadata and provenance

🧬 Execution Data Structure

Each Execution is stored as a self-contained directory:

execution/
├── auth/
│   └── jwt.json                 # Authentication info
├── metadata/
│   ├── method.json              # Method used
│   ├── request.json             # Inputs and parameters
│   ├── infrastructure.json      # Runtime platform
│   ├── instance.json            # Runtime container details
│   ├── jobStatus.json           # Lifecycle status
│   └── provo.xml                # Provenance metadata
├── outputs/
│   ├── output_file1.txt         # User-generated files
│   ├── stdout.log / stderr.log  # Execution logs
└── output.zip                   # Archived version of all outputs

🧪 How Executions Are Stored

Executions live in the user’s Workbench (temporary area) and can be:

Re-executed (same inputs)
Downloaded (as .zip)
Shared (via D4Science Workspace)
Archived to workspace under CCP/executions

They can also be imported back into the Workbench using:

A previously exported .zip
A Workspace reference

🚥 Execution Status

An Execution moves through these states:

CREATED → initial request accepted
SCHEDULED → assigned to an infrastructure
STARTED → deploy-script begins
EXECUTING → main script running
COMPLETED / FAILED → terminal state

Each state is logged in metadata_jobStatus.json.

📜 Metadata and Logging

CCP produces rich metadata for traceability:

provo.xml for reproducibility and FAIR principles
stdout / stderr logs
JSON summaries of method, infra, inputs, and runtime

You can inspect them from the execution view or download them.