Disaster Recovery as a Service (DRaaS), the ability to outsource an organization’s recovery site, typically to the cloud, is rapidly gaining in popularity. The executive team may assume the solution can do no wrong and encourage IT professionals to pursue it recklessly. Here are five things you need to make sure your boss understands about DRaaS.
1. The Cloud is Just another Data Center
The cloud is just another data center. IT professionals that work in them have to obey the same laws of physics an IT professional at a more traditional data center does. The advantage of the cloud is that it is available as the organization needs it, which makes it ideal for the disaster recovery use case. The organization only pays for the compute when it needs it during a disaster.
2. Storage Costs are Constant
While the organization only has to pay for cloud compute when it needs it, the storage cost is constant. It is necessary to seed data so it is available when compute is activated. That means it is also necessary to pay for storage costs. The constant cost of storage should be negligible if administrators manage it correctly. The only data the organization needs rapidly is the latest copy of data. The problem is most organizations look to use the backup process to seed the cloud, which means both the most recent data and all the old data are stored in the cloud, which raises the cloud storage costs.
3. Push Button Recovery Doesn’t Mean Instant Recovery
Most DRaaS vendors at some point will claim “recovery at the push of a button.” In most cases, this statement does not mean “instant” and IT executives may confuse the two. A push-button recovery means that the supplier of the solution has built in some automation to position the virtual instance of the application for recovery. After pushing that button, a series of background events need to take place and it may take minutes or even hours to recover.
4. Local Recovery is Still Critical
Most data centers won’t experience a disaster every day or even every year. The more common problems, disasters in their own right, are application failures, malware, server failures, and storage system failures. While DRaaS could stand-in for any of these failures, moving operations to the cloud for them is probably overkill. Instead, IT still needs the ability to recover on-premises. Make sure your boss knows you still need an on-site backup storage device and on-site backup software so in the event of a failure you don’t have to do all recoveries into the cloud.
5. There is still a Network Even after a Disaster
The fifth component that your boss needs to understand is that networking between users and the cloud after a disaster is going to take some work. When a disaster is declared, users will typically either be sent home, or sent to another office, or attempt to connect via a shared workspace. Making a secure connection will require some networking expertise, as well as planning and testing. A few vendors are coming out with Software Defined Networking (SDN) to aid IT planners in making these connections, but most require the organization to do all the work themselves.
6. Fail-back is Difficult
Failing back, no matter the location of the disaster recovery site, is potentially one of the hardest parts of DR planning. Typically, after a disaster, the original data center is in one of two states. Either the data center is completely wiped out, and all the data at the DRaaS location needs to be restored, or it can be 100% intact.
In the first case, the data center was destroyed, a complete recovery is required. While data could slowly be trickled down to the new data center, it is probably better to do a bulk transfer to get most of the data to the new data center quickly. The second case, the data center is fine, is surprisingly common. IT needs to figure out how to move data that was changed at the DR site back to the original data center. An incremental move like this is particularly difficult and requires planning as well as research during the selection process. Make sure your boss knows to ask how the original data center will be updated, if it survives the disaster.
While the cloud and services provided through the cloud can substantially improve an organization’s ability to respond to a crisis, they are not magic pills. DRaaS offers incredible potential to drive down the cost of disaster recovery planning. It still takes research, however, to find the right solution for the organization. It also takes careful implementation and testing to make sure that when disaster strikes, IT will know how to respond.