Skip to content

Latest commit

 

History

History
125 lines (90 loc) · 3.81 KB

README.md

File metadata and controls

125 lines (90 loc) · 3.81 KB
Raw

Simple Gateway for Measurements API

This project provides an API to insert and query measurement data in a DuckDB database. The API is built with FastAPI and supports data ingestion, custom SQL querying, and flexible data filtering for use with monitoring tools like Grafana.

Requirements

  • DuckDB: Ensure DuckDB is installed and accessible.
  • Python: Requires Python 3.7+ and fastapi, uvicorn, and duckdb libraries.
    Install dependencies:
    bash

pip install fastapi uvicorn duckdb

Setup

  1. Clone the repository and navigate to the project directory.
  2. Initialize a DuckDB database file named devices.duckdb:
    bash

duckdb devices.duckdb

Create the main measurements table within DuckDB:
sql
CREATE TABLE measurements (
bat_voltage DOUBLE,
dev_temp INTEGER,
device BIGINT,
device_sn VARCHAR,
device_urn VARCHAR,
env_temp INTEGER,
lnd_7128ec DOUBLE,
lnd_7318c DOUBLE,
lnd_7318u DOUBLE,
loc_country VARCHAR,
loc_lat DOUBLE,
loc_lon DOUBLE,
loc_name VARCHAR,
pms_pm02_5 DOUBLE,
when_captured TIMESTAMP,
received_at TIMESTAMP
);

Running the API

To start the FastAPI server:

bash
uvicorn grafana_duckdb_api:app --host 0.0.0.0 --port 8000

Or, for production use, start with Gunicorn and Uvicorn workers:

bash
gunicorn -w 4 -k uvicorn.workers.UvicornWorker grafana_duckdb_api:app --bind 0.0.0.0:8000

API Endpoints

POST /measurements

Ingest a new measurement into the DuckDB database.

  • Parameters:
    • api_key (required): A sample API key q1LKu7RQyxunnDW is used for testing purposes. Replace it as needed.
    • Payload: JSON with fields such as bat_voltage, dev_temp, device, etc. Empty string values are automatically converted to NULL by the API.
  • Example:
    bash

curl -X POST "http://localhost:8000/measurements?api_key=q1LKu7RQyxunnDW" \
-H "Content-Type: application/json" \
-d '{
"bat_voltage": "3.7",
"dev_temp": "25",
"device": "2830364905",
"device_sn": "SN12345",
"device_urn": "urn:dev:12345",
"env_temp": "30",
"lnd_7128ec": "60",
"lnd_7318c": "",
"lnd_7318u": "20",
"loc_country": "US",
"loc_lat": "37.7749",
"loc_lon": "-122.4194",
"loc_name": "San Francisco",
"pms_pm02_5": "12.5",
"when_captured": "2024-11-02T15:30:00"
}'

GET /query

Fetch data based on query filters for any of the supported fields. Useful for Grafana integration.

  • Parameters: Any field from the measurements table can be used as a filter, e.g., device, device_sn, when_captured, etc. Multiple device values can be specified with device=123&device=456.
  • Example:
    bash

curl -X GET "http://localhost:8000/query?device_sn=SN12345"

Example for Multiple Devices:
bash
curl -X GET "http://localhost:8000/query?device=123&device=456"

GET /status

A simple health check to verify that the API is running.

Notes

  • Dynamic Column Addition: The API automatically adds any new columns to the measurements table if new fields are received in the JSON payload.
  • Handling Nulls: Empty string values are converted to NULL before insertion into DuckDB, ensuring data integrity and allowing accurate querying.

Grafana Integration

For Grafana, use the Infinity datasource to connect to the API:

  1. Set up a Yesoreyeram Infinity datasource.
  2. Set the URL to http://localhost:8000/query and specify query parameters using JSONPath or direct query fields.
  3. Configure queries as needed for custom visualization.