Clarify the some arguments from ray.serve.deployment

From the documentation, I saw two arguments in ray.serve.deployment:

*graceful_shutdown_wait_loop_s: Union[ray.serve._private.utils.DEFAULT,float] = DEFAULT.VALUE* , 
*graceful_shutdown_timeout_s: Union[ray.serve._private.utils.DEFAULT,float] = DEFAULT.VALUE* ,

Could I get some clarifications about what they are and how to use them?

Hi @Frank, thanks for asking!

From the deployment schema documentation:

  • graceful_shutdown_wait_loop_s: Duration that deployment replicas will wait until there is no more work to be done before shutting down. Uses a default if None.
  • graceful_shutdown_timeout_s: Serve controller waits for this duration before forcefully killing the replica for shutdown. Uses a default if None.

In the future, please post questions about Ray Serve in the #ray-air:ray-serve category for better visibility and quicker responses.