Delete all cortex-created AWS resources when deleting a cluster

[deliahu]

Notes

Consider adding a flag which would allow the user to preserve cluster state. If so, add docs describing how to delete the resources later. It could be best if cortex cluster down deleted them even if the cluster was already deleted previously (without deleting the resources).

Depending on the above, update cluster uninstall docs accordingly.

• log groups (can potentially only be a single one, in which case the delete command will be simpler)
• rename keep-volumes to keep-aws-resources?
• make sure that deletes of resources are independent and best effort
• report errors to sentry at the end of the program's execution

• make sure that deletes of resources are independent and best effort
  • fully functional cluster
    • delete everything sequentially and log everything
  • partially deleted cluster
    • delete everything that hasn't been deleted for the cluster
  • completely deleted cluster
    • same as for the above (already deleted x,y,z).