Spike | Have admission controller fault tolerant

[MostafaMegahid]

The dependency on having the admission controller working all the time on the clusters is increasing, especially with the tenancy work being dependant on having the admission controller.

Objective

Explore options to have the admission controller flow fault tolerant.

Remarks

• The flow of the audit is becoming risky to have running within the same pod as the admission controller, specially with enabling the users to configure the audit frequency
• We should identify other scenarios that might break the agent, and subsequently break the admission controller flow, then propose solutions to the top riskiest scenarios.