A lightweight service observability platform for uptime checks, latency tracking, incident history and Prometheus metrics.
DevPulse monitors HTTP services, stores check results, evaluates alert rules and exposes application metrics for Prometheus. It includes a REST API and a small dashboard-ready data model.
Status: functional portfolio MVP. Checks are triggered through the API; a background scheduler, authentication and multi-tenant isolation are future work.
- Register and manage HTTP service targets
- Trigger HTTP health checks with configurable timeouts
- Persist status code, latency, availability and error details
- Evaluate latency and availability alert thresholds
- Maintain an incident/event history
- Expose Prometheus metrics at
/metrics - Structured health endpoint and OpenAPI docs
- Docker Compose, PostgreSQL and automated tests
+-----------------------+
| Dashboard |
+-----------+-----------+
|
v
+-----------------------+
| FastAPI API |
+----+-------------+----+
| |
Target registry Check runner
|
v
HTTP services
|
v
+-----------------------+
| Check results + |
| alert evaluation |
+-----------+-----------+
|
+---------+---------+
| |
v v
PostgreSQL Prometheus
/metrics
Python 3.12 · FastAPI · SQLAlchemy 2 · PostgreSQL · httpx · Prometheus Client · pytest · Docker
git clone https://github.com/asunavlr/dev-pulse.git
cd dev-pulse
cp .env.example .env
docker compose up --build- API: http://localhost:8000
- Swagger: http://localhost:8000/docs
- Prometheus metrics: http://localhost:8000/metrics
- Health: http://localhost:8000/health
python -m venv .venv
# Windows: .venv\\Scripts\\Activate.ps1
source .venv/bin/activate
pip install -e ".[dev]"
uvicorn devpulse.main:app --reload
pytest| Method | Endpoint | Description |
|---|---|---|
| GET | /health |
Service health |
| GET | /metrics |
Prometheus exposition format |
| POST | /targets |
Register a monitored HTTP target |
| GET | /targets |
List targets |
| GET | /targets/{id} |
Target details and latest check |
| DELETE | /targets/{id} |
Delete target and its history |
| POST | /targets/{id}/check |
Execute an HTTP check |
| GET | /targets/{id}/checks |
Recent check history |
| GET | /incidents |
List alert events |
curl -X POST http://localhost:8000/targets \
-H "Content-Type: application/json" \
-d '{"name":"Example API","url":"https://httpbin.org/status/200","timeout_seconds":5,"latency_threshold_ms":800}'curl -X POST http://localhost:8000/targets/1/checkA check records the HTTP status, duration and outcome. A failed request or a latency threshold breach creates an incident event. Repeated checks can resolve an active incident when the service recovers.
DevPulse exports:
devpulse_checks_total{target,status}devpulse_check_duration_seconds{target}devpulse_target_up{target}
Example Prometheus scrape configuration:
scrape_configs:
- job_name: devpulse
static_configs:
- targets: ["devpulse:8000"]| Variable | Default | Description |
|---|---|---|
DATABASE_URL |
SQLite file | SQLAlchemy URL |
DEFAULT_TIMEOUT_SECONDS |
5 |
Default outbound HTTP timeout |
MAX_TIMEOUT_SECONDS |
30 |
Maximum target timeout |
CHECK_HISTORY_LIMIT |
100 |
Maximum returned history rows |
- Target URLs are user-supplied. This MVP is intended for trusted, local/demo environments; do not expose target registration to untrusted users.
- Production deployment should enforce outbound network allowlists and block loopback, private, link-local and cloud metadata addresses to prevent SSRF.
- Authentication, RBAC, tenant isolation and audit controls are not implemented yet.
- Checks are manually triggered; no scheduler or distributed worker is included yet.
- Background scheduler and distributed workers
- OpenTelemetry traces and context propagation
- Alert routing (email, Slack, webhooks)
- SLOs, error budgets and burn-rate alerts
- Multi-region probes
- Authentication, RBAC and tenant isolation
- Dashboard with live charts and incident timeline
- Network egress policy and SSRF protection
MIT