Open
Description
A user reported that they are running sky launch
but they find sky launch
on existing cluster is very slow and the expect behavior is that:
- if cluster does not exist, provision the cluster and run the job
- if the cluster exists, run the job only (like exec), and skip all those time consuming steps, including skypilot runtime setup, waiting for ssh, and user setup.
Two ways to achieve this:
- Make the
sky launch
super fast on an existing cluster by caching the current state of a cluster and only re-setup the cluster when the runtime is stale. - add an option to automatically use
sky.exec
whensky launch
is run on an existing cluster.