Depends if you have the same subnet network in both sites (think VXLAN between sites) or if you have completely separate networks. If you want immediate failover, you'll have to either deal with Always On Availability Groups which is duplicating data on each cluster node or have a geo-stretch windows failover cluster with a single SQL Server Failover Cluster Instance installed that has IP Addresses defined for both networks (whether the same actual subnet or different) and handle the data replication at a SAN level. With this, you would need to find a solution to automate the SAN failover. There are other options with VMWare and Hyper-V virtualization and replication of VM's between datacenters as well. You have a lot of options that can depend on your budget and access to technology. You'll also want to determine your recovery point objective (RPO) and recovery time objective (RTO) and ensure you understand any potential data loss if you're not using synchronous data replication. Don't forget, each of these solutions require different SQL Server Licensing requirements which can be $ignificant.