feat: system controller improvements after migration

[raelga]

Why?

After deploying system onto OCP, several improvements to the 3scale SaaS Operator had been discussed.

How?

By adding all the required functionalities to the system workload (and the others if required).

Requirements

• [x] Bump Grace Period from 30s to 60s
  • [x] System components
  • [x] Other workloads?
• [x] Change rollout behaviour to scale using a custom % of the pods
  • [x] System components
  • [x] Other workloads?
• [x] Change HPA behaviour to scale using a custom % of the pods
  • [x] System components
  • [x] Other workloads?
• [x] Change PDB to allow rollout behaviour if required
  • [x] System components
  • [x] Other workloads?
• [x] Allow custom image for the system console
• [x] Add twemproxy sidecar for the system console

Acceptance Criteria

• [x] Bump Grace Period from 30s to 60s
  • [x] System components
  • [x] Other workloads?
• [x] Change rollout behaviour to scale using a custom % of the pods
  • [x] System components
  • [x] Other workloads?
• [x] Change HPA behaviour to scale using a custom % of the pods
  • [x] System components
  • [x] Other workloads?
• [x] Change PDB to allow rollout behaviour if required
  • [x] System components
  • [x] Other workloads?
• [x] Allow custom image for the system console
• [x] Twemproxy sidecar for the system console

Notes

/kind feature /priority important-soon /assign