Huge repository validation

[tusmester]

Validate that the repository can serve various kinds of requests (reads and writes) even with a huge repository containing many content items.

Create a huge database

Create a package or mechanism that can repeatedly fill any repository with test data. Acceptance criteria:

• use sn IO if possible, instead of a custom import
• should be a self-supporting package containing all the import items (necessary CTDs, users, permissions, etc.)
• test CTDs: should be business-focused, as real life as possible, based on previous projects
• determine the number of items per content type and their relations

The benchmark

Test the repository with automated requests.

• repeatable, scripted tests
• many reads, fewer writes
• first: directly using the OData REST API, later add a web API service in the middle
• GOAL: determine the threshold: how many requests can we serve in a stable way?
• document the results