olly - Fotolia
It depends, but it is likely that you would use virtual server technology to transition the workload back to a machine in your own data center. But if you go down the path of building a failover option in the cloud, you have to do the work to make sure that the application runs in the cloud.
Once you have got that set up, you really ought to ask yourself why even fail back at all. If you know this application works in the cloud and that it can serve customers and stakeholders from the cloud, then what's the point of failing it back or maintaining the physical data center presence?
Would you be better off simply freeing up that physical server capacity and maybe using it for stuff that's not compatible with the cloud, whether it's a technology compatibility issue or maybe a regulatory issue?
For you to run a DR version of an application of the cloud means you have to be willing to have that data live in the cloud and have the workload serviced by the cloud. The performance has to be OK, and security, privacy, etc. have to be acceptable.
Once you've made those decisions, then it doesn't make a ton of sense to spend of lot of effort bringing it back and keeping it primary. What makes more sense is to simply migrate that whole application to the cloud. DR could be a good interim step to be able to test it out and decide if it will work for you.
How cloud DR service stacks up with DR testing
Are cloud DR services secure?
Pros and cons of migrating DR services to the cloud
Dig Deeper on Disaster recovery storage
Related Q&A from Andrew Reichman
When evaluating cloud disaster recovery, you will need to think about what your environment looks like now and what it will look like in the next few... Continue Reading
Cloud-based disaster recovery sounds almost too good to be true. What are some challenges to look out for? Continue Reading
Learn some of the bandwidth challenges associated with backing up data to the cloud. Continue Reading