One of the ways the Community will access the Commons Impact Metrics data is via public API (AQS2).
A first step to achieve this goal is to design the endpoints of the Commons Impact Metrics AQS service.
We should do this after we reviewed the feedback from the Commuinity in T358688.
We should take into account that AQS will query Cassandra or Druid,
and also design the corresponding source datasets.
Tasks:
- Review the different documents to have a good idea of what the users need/want, in which format, granularity and dimensions they would like to receive the data. Also it is important that the endpoints are compatible with the existing data (if an endpoint requires small modifications of the data model, that's fine).
- Data model and implementation plan
- Commons Analytics Product Needs research
- Feedback from the Community on the prototype data model (no link yet T358688).
- Design a list of endpoints. Each endpoint should respond to 1 question (or a family of similar questions) from the users. Each endpoint design should specify:
- URL format
- Metric name
- Granularity
- List of dimensions (for each dimension: type, format, possible values, semantics)
- List of metric values
- Format of the JSON results
- Design the list of Cassandra/Druid datasources needed by the endpoints. This will be the contract that other tasks will use to make sure the pipeline works.
- table keyspaces and names
- Keys and values
- Format of the blobs if any
- Create table statements
- ...?
Deliverables:
- A document with the design of the endpoints and the datasources