Skip to content

Latest commit

 

History

History
156 lines (129 loc) · 4.05 KB

Python_README.md

File metadata and controls

156 lines (129 loc) · 4.05 KB
Raw

Python instructions

  1. Sensor App
  2. Sensor API

2. Sensor App

You will be running the following steps from the command line.

Note: since the code uses the python-dotenv package, sourcing the dot-env file is not even necessary for Python.

Setup

First place yourself to the proper folder and install requirements.

cd python
pip install -r requirements.txt

The core module is db_connection.py which provides a (Singleton) Cassandra session for use by the rest of the code. Take a minute to inspect that.

Run stuff

💻 Run connectivity test:

python ex_00_connectivity.py
Show expected result 
$> python ex_00_connectivity.py
[get_session] Creating session
    ** Connected to cluster 'cndb' at data center 'eu-west-1' **
[shutdown_driver] Closing connection

💻 Run query Q3 as standalone exercise:

python ex_01_query_Q3.py volcano-net
Show expected result 
$> python ex_01_query_Q3.py volcano-net
[get_session] Creating session
    ** Querying sensors for network 'volcano-net' ...
      - Sensor 's2001'  (LAT=+44.46, LON=-110.83): accuracy = high, sensitivity = medium
      - Sensor 's2002'  (LAT=+44.46, LON=-110.83): accuracy = high, sensitivity = medium
[shutdown_driver] Closing connection

💻 Try to run the improved form of the same exercise, which uses prepared statements:

python ex_01B_query_Q3.py volcano-net

If you plan to run the same CQL statements over and over in your driver-based application (possibly with variable arguments), you should always employ prepared statements as they improve the performance by reducing the overhead. Thus this version of query Q3 is to be preferred.

3. Sensor API

💻 Start the API:

uvicorn api:app
Show expected result 
$> uvicorn api:app
INFO:     Started server process [68610]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)

💻 With the API running, in the other shell try to call the "Q3" endpoint (GET):

curl -s localhost:8000/sensors_by_network/volcano-net | jq
Show expected result 
$> curl -s localhost:8000/sensors_by_network/volcano-net | jq
[
    {
        "characteristics": {
            "accuracy": "high",
            "sensitivity": "medium"
        },
        "latitude": 44.460321,
        "longitude": -110.828151,
        "network": "volcano-net",
        "sensor": "s2001"
    },
    {
        "characteristics": {
            "accuracy": "high",
            "sensitivity": "medium"
        },
        "latitude": 44.463195,
        "longitude": -110.830124,
        "network": "volcano-net",
        "sensor": "s2002"
    }
]

This endpoint is a GET and its parameter is a path component in the URL. Try to find, in the API code, where the URL path is parsed to obtain the network name.

💻 With the API running, in the other shell try to call the "Q4" endpoint (POST):

curl -s -XPOST localhost:8000/measurements_by_sensor_date \
    -d '{"sensor":"s1001", "date":"2020-07-04"}' \
    -H 'Content-Type: application/json' | jq
Show expected result 
$ curl -s -XPOST localhost:8000/measurements_by_sensor_date \
>     -d '{"sensor":"s1001", "date":"2020-07-04"}' \
>     -H 'Content-Type: application/json' | jq
[
  {
    "timestamp": "2020-07-04T12:59:59",
    "value": 98
  },
  {
    "timestamp": "2020-07-04T12:00:01",
    "value": 97
  },
  {
    "timestamp": "2020-07-04T00:59:59",
    "value": 79
  },
  {
    "timestamp": "2020-07-04T00:00:01",
    "value": 80
  }
]

In this case, the parameters are passed as POST payload: you can check, in the API code, the way these are parsed and used within the endpoint function body. This makes use of a pydantic model and the highly-automated dependency management facilities offered by FastAPI.