optimize scalecube - serverLifePeriodMS option

[katzIdo]

General description

every microservice container can act as a seed for another microservice container in our distributed env.

it results in inflation of unnecessary listeners which are added to the project.

every microservice that acts as a seed creates a "server-side" logic which means using a global listener that checks if it's been called.

my suggestion is to add to microservice API another property which called "serverLifePeriodMS: number"

if (serverLifePeriodMS === undefined) {
  // server listener will live forever
} else {
  // server will stop listen after  serverLifePeriodMS is done
}

this option is good when we creating an empty microservice that is used only to get proxies of other microservice containers, we can pass serverLifePeriodMS = 0.

or if we are running on the browser and we know that our services will load after x period of time, then we can set our serverLifePeriodMS to be 2*x.

this option is valid for the browser because it is unusual for service to be "destroyed" in the browser.

this is another reason why cluster-browser does not implement a "keep-alive" check.

API

Design proposition

A clear and concise description of the design proposition.

• Is backward compatible

• Adds new API

Pull Request

[Link]

New API

Code snippet

export interface MicroserviceOptions {
  // ... all previus methods unchanged
 serverLifePeriodMS : number
}

Ready for implementation

• External dependencies have been resolved

• API has been approved

• Test cases have been prepared

• Discussed with Technical lead

Definition of Done

• Developer review

• QA review

• All tests are implemented

• Documentation