Timeseries

This is an coding practice repository for a timeseries API using python, angular, mqtt and MongoDB.

      

Requirements

• Pyenv
• Pyenv virtualenv
• GNU make

Development

Run all tests with docker compose:

make test

Spin up development environment:

make compose-up

Query String Filters

API endpoint with query string filters deep-object like:

?timestamp[gte]=2021-01-01T00:00:00&timestamp[lte]=2021-01-05T00:00:00&metadata.area=kitchen&limit=100"

will become mongo query:

{
    "timestamp": { $gte: ISODate("2021-01-01T00:00:00"), $lte: ISODate("2021-01-05T00:00:00") },
    "metadata": { "area": "kitchen" },
    "limit": 100
}

CI

Used github actions to define CI pipelines:

Preview

Default endpoint result:

Limits to 100 results:

Angular Custom Dashboard:

Plotly Dashboard:

Filtering by query string example:

Dependencies

TODO

• API pagination offset (done)
• Pyparsing query validation (done)
• check usage of mongo engine ORM with timeseries collection
• kube-linter does not support kustomize, pre-render manifests before running linter
• add UI tests (cypress)
• increase test coverage (done)
• instrumentation, tracing, metrics and monitoring
• terraform deployment