CloudHSM cluster always ends up in degraded state

I’m following Ryan’s instructions here to create CloudHSM but each time I end up with a "Degraded" cluster. AWS just tells me to delete and recreate with no error message or explanation. Anyone run into this?