A production application must survive a single compute hardware maintenance event within an availability domain. Which placement should the team consider for its redundant instances?

1. One boot volume with no redundant application instance
2. The same instance with two display names
3. Different fault domains
4. Two compartments on the same physical failure domain only