As a user I have efficient collections endpoint

[pulpbot]

Author: @fao89 (fao89)

Redmine Issue: 8262, https://pulp.plan.io/issues/8262

---

Currently, the collections endpoint provides data from multiple tables which can harm the performance, as the query is done for every request.

We need to pre-compute the query to improve the performance

Context: https://github.com/pulp/pulp_ansible/pull/469/files#r573247512