Does that automatically respect the request_runtime of the jobs because it may tinker with the uptime value of the host ?
I don't understand this question. AFAIK, "request_runtime" isn't a thing -- at least, it doesn't appear to be in the manual. I'm also not sure what the "uptime value" of the host is in the context -- the host's uptime (time since boot) shouldn't change any as a result of running or not-running jobs.
-- ToddM