Proposes change to rate_limits_spec

What does this MR do and why?

This change asserts on too many requests after rate limit has been reached for an authenticated user.

It will depend on an omnibus configuration with 0 cache seconds for application settings

Screenshots or screen recordings

N/A

How to set up and validate locally

1. in gitlab.yml set the application settings cache seconds:

gitlab:
  application_settings_cache_seconds: 0

2. restart local rails app
3. in gitlab/qa folder run bin/qa Test::Instance::All http://localhost:3000 -- -t rate_limits

MR acceptance checklist

This checklist encourages us to confirm any changes have been analyzed to reduce risks in quality, performance, reliability, security, and maintainability.

• I have evaluated the MR acceptance checklist for this MR.