A critical service must tolerate losing an entire building. Which design directly addresses that failure scope?

1. Create a second virtual machine on the same host
2. Deploy redundant service capacity in a separate failure location
3. Add a second power supply in the same server
4. Use two disks in the same desktop computer