Mock High Value Cloud Providers

Inizia dai documenti di discovery di GCP collegati, dalla documentazione REST di Google Directory e Drive e dalle risorse Kubernetes OpenAPI; nell’issue non sono identificati file del repository né test. Il lavoro sarà considerato completato quando saranno disponibili mock Flask eseguibili separatamente, con risposte basate su Jinja, porte arbitrarie, esempi fedeli allo schema, evidenze con curl e una guida da zero che copra le interfacce richieste.

Mock High Value Cloud Providers

Background

We want python3 flask mock servers to mock a large breadth of cloud services. These should be supplied as a set of separate applications per "service" offering (eg: GCP Compute API, GCP Big Query API, Google Cloud Storage API...). Response bodies should leverage jinja templates so that changing the response or adapting to different input is purely a function of altering templates... no code changes to change responses. Each flask app should be runnable on an arbitrary port, no port hard coding. The response bodies should conform to the schemas of actual responses and should resemble vendor examples and/or schema constraints, where possible. For example, the GCP Coumpute instances list endpoint should return somoe object that is faithful to the response schema in the endpoint doco.

The following are required to be prepared:

1. The Google Cloud Platform (GCP) API for all GCP services. For each of these, it is helpful to provide a mock for all supported versions. Each one of these artifacts should be a distinct flask app.
2. The Google directory and drive APIs. (These are separate to GCP).
3. The entire kubernetes API. This can be one flask app or more than one; as long as it works.

Helpful resources:

• GCP discovery documents are available for all cloud services using google's discovery service.
• The google directory API doco.
• The google drive API doco.
• The kuberetes API can be downloaded in various formats; please see this doco.

Deliverables

1. python3 flask mock servers for all above mentioned interfaces.
2. Codebase used to generate all output materials, as a private github repository. MIT Licensed.
3. Evidence of responses reflecting examples, eg: curl. 2 or 3 examples is fine.
4. Simple human readable from scratch walkthrough on how to start a server and test manually with curl.

Budget

1,000 AUD.

Bonuses

1. Extra 2,000 AUD if AWS, Azure (including Entra ID), GitHub, Okta, DataBricks, Confluent are also delivered.