skypilot-org / skypilot

SkyPilot: Run AI and batch jobs on any infra (Kubernetes or 12+ clouds). Get unified execution, cost savings, and high GPU availability via a simple interface.
https://skypilot.readthedocs.io
Apache License 2.0
6.48k stars 462 forks source link

UX: when launching, print currently in-effect config.yaml #1860

Closed concretevitamin closed 1 year ago

concretevitamin commented 1 year ago

Got tripped up because my config.yaml has private VPC/jump server setting that I forgot about. The launch didn't go through, timing out only after 10+ minutes.

Michaelvll commented 1 year ago

Also, can we do a pre-check to see if the proxycommand is correct, i.e. if the jump server can be connected, before we start launching the cluster?

concretevitamin commented 1 year ago

One problem is ssh_proxy_command itself is not necessarily a well-formed SSH command:

ssh_proxy_command: ssh -W %h:%p -i ~/.ssh/sky-key -o StrictHostKeyChecking=no ec2-user@1.2.3.4

Directly running the above will lead to

Bad stdio forwarding specification '%h:%p'

We could try to match <usr>@<ip>, but in general users may pass in something like ./my_proxy.sh. I think we can add some printing of currently in-use VPC/SSH settings first, and defer the best-effort jump server ping for later?

github-actions[bot] commented 1 year ago

This issue is stale because it has been open 120 days with no activity. Remove stale label or comment or this will be closed in 10 days.

github-actions[bot] commented 1 year ago

This issue was closed because it has been stalled for 10 days with no activity.