This message was deleted.

Things may have changed now, but in my past experience, changing cloud providers post cluster creation results in some very odd behavior. Initially it may work fine but for me things took bad turn during node reboots.