Epic Goal

Support High Availability support easily from the Operator. This means, per components (Pipeline, …) :

• Configuring webhook HPA and replicas
• Configuring controller's HA setup (buckets, …) and replicas

See https://github.com/tektoncd/pipeline/issues/6976 as well

Why is this important?

We need to make sure users can scale the components elements in order to provide better performance and high availability (in case of node failure, …).

This can have a direct impact on existing customers that are running into performance issues.

Scenarios

This is for Pipeline components, but most likely applies to any.

• I want to configure the controller to be higly available in case of a node failure — aka if a node fails I don't want a service interruption. But I only want one controller at a time working on objects.
• I want to spread the load between multiple controller that are running in different nodes (to also spread the "load" between multiple nodes)

Acceptance Criteria (Mandatory)

• It should be easily to configure different scenario for HA (fail-over, spread the load)
• It should be doable per component
• One component's configuration shouldn't affect the other (not the case today due to all components reading the same configmap)

Dependencies (internal and external)

N/A

Previous Work (Optional):

https://github.com/tektoncd/operator/pull/1458

Open questions::

N/A

Done Checklist

• Acceptance criteria are met
• Non-functional properties of the Feature have been validated (such as performance, resource, UX, security or privacy aspects)
• User Journey automation is delivered
• Support and SRE teams are provided with enough skills to support the feature in production environment