It may happen that a stack build can fail, where the failures could be caused by temporary network errors, like:

Error: endpoint/kloud: Failed to dial the following kites:

  * my-server: no kites availabile (id="///////92873000-029a-0000-aa1f-63506aa70000", url="", state="kontrol")

In order to ease troubleshooting the reason, do:

• generate a ssh key for each stack credential and inject it with cloud-init for all vms for all providers (allow for opt-out from that behaviour)
• extend the error message with a vm ip
• allow for reading remote klient logs with kd logs <vm>
  • research if it's possible to read vm details if stack is in failed state (klient/remote /cc @ppknap)