[Impl] implement internal Job runner

[vankichi]

Is your feature request related to a problem? Please describe.:

The current runner effectively manages long-term workloads like API servers, launching, terminating, and handling signals for proper shutdown.
However, handling short-term workloads like Jobs requires more complex management, as processes must self-terminate by sending SIGTERM when their use cases conclude.

Describe the solution you'd like:

Design and implement a runner for the Job, replacing the current Job runner.

Describe alternatives you've considered:

For short-term workloads, we want to stop the workload at any time (such as when the use case ends), so we design and implement a mechanism to achieve this.