Description

Here is a simple metrics uploader importer.py (saves test_data.csv to prepared DB) and the service to read the imported data in simple HTML format.

Requirements

1. Postgres DB
2. Python 3.7
3. pipenv

IMPORTER

Its a simple CLI application which allows you to save test_data.csv to a DB. Be aware, you need to setup DB first, before applying CSV to it.

How To

1. run python -m pipenv install to create a virtual environment and install dependencies
2. run sh resources/init.sh to create DB and all needed tables
3. run python importer.py --csv resources/task_data.csv --url postgresql://<user>:<password>@<host>/<db_name>

NOTE: use your own username and password for DB (in init.sh and --url), mine are default from Docker Postgres image

To check that data was really imported the DB:

1. run psql -h <host> -U <user> and type password after to connect to Postgres
2. run \c <db_name> to connect to DB
3. run SELECT * FROM metrics;

SERVICE

This is a Flask-RestX simple service to get data from DB and return it as simple HTML. Each request to this service is logged separately to table logs in DB.

How to

1. make sure you run python -m pipenv install to install all dependencies and virtual environment
2. make sure that DB is created (see resources/init.sh) and populated (see Importer section)
3. make sure your DB credentials are valid in service/config.py
4. run python service/app.py to run an application

To check the output call with CURL or in Browse http://127.0.0.1:5000/api/v1/metrics

Also the service supports query param limit, it will return exact amount of rows from DB what was specified. For example http://127.0.0.1:5000/api/v1/metrics?limit=1 If this parameter is not specified, you will get the full output.

To check that request was logged run SQL query in DB (see Importer section): SELECT * FROM logs;

Testing

To run unit tests use:

python -m unittest discover -s tests/ -v

Comments

1. Its better to use env variables in init.sh instead of hardcoding them
2. Importer is not ideal, it would nice to have it more flexible, to be able upload different CSVs, and with better validation
3. There is no unit-tests for Importer, I understand it is important to have
4. TODO: get output by timestamp
5. Instead I added limit parameter