deploy-simulator

CI/CD deployment simulator.

Setup

From the project root:

python3 -m venv .venv
source .venv/bin/activate   # Windows: .venv\Scripts\activate
pip install -r requirements.txt

Running the app

The API and frontend run on a single server — there is no separate frontend process. FastAPI serves the REST API and static files (App/static/index.html, App/static/styles.css) together.

From the project root, with the virtual environment activated:

uvicorn App.main:app --reload

What	URL
Frontend UI	http://127.0.0.1:8000/
API docs (Swagger)	http://127.0.0.1:8000/docs
Deployments API	http://127.0.0.1:8000/deployments

Using the UI

1. Open http://127.0.0.1:8000/ in your browser.
2. Toggle Deployment generation on to call POST /deployments/start (adds a deployment every 10 seconds).
3. Toggle it off to call POST /deployments/stop.
4. The deployments table refreshes automatically every 5 seconds via GET /deployments.

Stop the server with Ctrl+C in the terminal.

Optional flags

# Custom host/port
uvicorn App.main:app --host 0.0.0.0 --port 8080 --reload

Data model

App/models.py defines a Pydantic Deployment record:

Field	Type	Notes
id	int	Unique identifier, >= 1
service	literal	billing-api, auth-service, notifications, frontend-web
status	literal	running, success, fail, rolled-back (running has duration 0 until completed)
duration	float	Duration in seconds
timestamp	str	ISO 8601 timestamp
commit_sha	str	Git commit SHA (7–40 chars)

Deployment operations

App/deployment_ops.py defines DeploymentOps, an in-memory store backed by a Queue[Deployment]:

• add_deployment(deployment) — add a running deployment (status must be running, duration 0)
• complete_deployment(deployment_id, status, timestamp) — finalize a running deployment; duration is computed from timestamps
• read_deployments(service=None, status=None) — list deployments, optionally filtered
• read_running_deployments(service=None) — list in-flight deployments
• read_deployment(deployment_id) — fetch one deployment by id

Metrics

App/metric.py defines MetricOps for analytics over completed deployments:

• record_completed(deployment) — update per-service stats, status counts, and anomaly detection (called by DeploymentOps on completion)
• read_anomalies() — list anomalous deployments
• success_rate() — success count / total × 100
• p95_duration() — 95th percentile duration per service

Dummy data generation

App/dummy_generator.py defines DummyGenerator for synthetic deployment records:

• create_dummy_deployment() — build one random running Deployment and return it as a dict
• start_deployments(ops) — background thread starts a running deployment every 10 seconds and completes them after a simulated duration
• stop_deployments() — stop the background generator

API

Method	Path	Description
GET	/deployments	List deployments; optional service, status query filters
GET	/deployments/latest	Latest completed deployments (up to 10)
GET	/deployments/running	All in-flight running deployments; optional service filter
GET	/deployments/{id}	Get one deployment by id (id >= 1)
PATCH	/deployments/{id}	Complete a running deployment (status, timestamp in body)
GET	/p95	P95 duration per service (or not enough deployments if fewer than 5)
GET	/anomalies	List of anomalous deployments
GET	/success-rate	Success rate percentage { "status": "success", "value": 95.0 }
POST	/deployments/start	Start dummy deployment generation
POST	/deployments/stop	Stop dummy deployment generation

Errors return a structured body: { "error", "message", "code", "details" }.

Project layout

• App/ — application code (main.py, models, static frontend)
• App/static/ — index.html, styles.css
• Root — README.md, .gitignore, requirements.txt