Priority

Priority bands specify the priority of a request for eventbus quota enforcement.

At the time we setup the Client, we grant them per-role quotas in one or more priorities, client can adjust these quotas by contacting eventbus team.

At the time client fires workflow, they should send the priority information along with the request either by set the RequestQoS (evebtbus map the value to the one has same name in Priority proto), or set the priority field in the rpc request. When both setted, eventbus ignore the RequestQoS for our own quota enforcement (note that this won't impact the the rpcs the workflow makes during execution, aka. RequestQoS instead of priority in the rpc request will be passed to the rpcs the workflow makes).

Enums
UNSPCIFIED Unspecified
SHEDDABLE Frequent partial and occasional full unavailability is expected and not pageable. * Requests to this band will be shed before all other requests. * This is the default for async calls sent from batch jobs.
SHEDDABLE_PLUS Partial unavailability is expected and is not necessarily pageable. * Requests to this band will be shed before any critical traffic. * This is the default for async calls sent from production jobs.
CRITICAL Any outage is a pageable event. * During a production outage requests in this band will only be shed before CRITICAL_PLUS. * This is the default for sync calls sent from production jobs.
CRITICAL_PLUS Any outage is a pageable event. * The guideline is for < 10% of requests to a service to be in this band. * During a production outage requests in this band will be prioritized above all others. * Opt-in to CRITICAL_PLUS when your workflow triggers by human.