Support updating all BE pods in parallel in the same time.

[dengliu]

Describe the problem

Currently it is the default rolling update, and the pods are updated one after another. This is extremely slow. and during pod restarts, ongoing queries/ingestion could be disrupted.
We need an option to config the update to be rolling out all 10 pods in parallel to reduce the instability window. so we can plan a narrow downtime window to complete the rollout quickly.

Describe the solution you'd like

A clear and concise description of what you want to happen.

Additional context

Add any other context or screenshots about the enhancement here.

[yandongxiao]

please see #594