Disaster recovery as a service (DRaaS)
Also known as: DRaaS, disaster recovery as a service, cloud DR, failover service
Disaster recovery as a service (DRaaS) continuously replicates your servers and data to a cloud provider so that, if your systems fail, you can 'fail over' and run them from the cloud almost immediately.
It delivers fast, tested recovery from major outages without the cost of building and maintaining a second data centre.
Disaster recovery as a service (DRaaS), explained at three levels
At home
Imagine your restaurant has a fully stocked twin kitchen across town, kept up to date every few minutes and ready to open the moment yours has a fire. DRaaS is that twin kitchen for your computer systems. Your servers are continuously copied to a provider's cloud, and if your office systems fail, the copies are switched on there so staff can keep working.
At work
Disaster recovery as a service (DRaaS) is a subscription where a provider continuously replicates your servers to its cloud and can run them there if your own systems fail. For a business, it replaces the cost of a second data centre with a monthly fee and cuts recovery time from days of rebuilding to minutes or hours of switching over.
At home
DRaaS works in three phases. Replication sends changes from your servers to the provider every few minutes or seconds, so the cloud copy stays current. Failover happens during a disaster: the provider boots your servers as virtual machines in its data centre and reconnects users. Failback happens after repairs: changes made while running in the cloud are copied back to your own hardware, and you switch home.
At work
A business chooses DRaaS based on two targets from its business impact analysis: the recovery time objective (RTO), how long systems can be down, and the recovery point objective (RPO), how much recent data it can lose. Plain cloud backup stores data but needs new hardware to restore onto. DRaaS adds standby computing, networking, and orchestration so the systems actually run. Providers include Datto, Zerto, and Veeam Cloud Connect partners.
At home
The hard part of DRaaS isn't copying data, it's bringing a whole environment up in the right order. A runbook defines the boot sequence, so the domain controller starts before the database, and the database before the application. It also maps networking and DNS so users can reach the cloud copy. While test failovers take effort, in my view a DR plan that hasn't been failed over at least once is unproven.
At work
Selecting DRaaS means matching RPO and RTO tiers to cost, confirming orchestration handles application dependencies, and testing network cutover through VPN, DNS changes, or a remote desktop gateway. Ask whether the provider runs failover for you or leaves it self-service, where replicas are hosted, and whether that meets Canadian data residency needs. Common failures include untested runbooks, licensing that doesn't cover cloud failover, and no clear authority to declare a disaster.
Real-world examples
At home
Lena's design studio depended on one server holding all client files and its project software. Her MSP set up DRaaS so the server replicates to the cloud. During a hardware failure, they failed over to the cloud copy and the team kept working within the hour.
At work
A fire made Carlos's office unusable. Because his critical systems were protected by DRaaS, the MSP failed everything over to the cloud, staff connected from home, and the business kept operating while the office was rebuilt, turning a potential closure into a temporary relocation.
How it compares
| feature | Cloud backup | DRaaS |
|---|---|---|
| What it protects | Copies of data/files | Running servers and systems |
| Recovery time | Hours to days (rebuild) | Minutes to hours (fail over) |
| Can run workloads | No (storage only) | Yes (in the cloud) |
| Cost | Lower | Higher (replication + compute) |
| Best for | Data retention, small footprints | Systems that can't be down long |
For analysts
DRaaS is worth it when downtime is expensive. Calculate your cost per hour of being unable to operate. If a day-long outage would seriously harm the business, the subscription is usually far cheaper than the downtime it prevents.
History & origin
Traditional disaster recovery meant buying and maintaining duplicate hardware at a second site, expensive and often neglected. Cloud computing made it possible to replicate systems to a provider and spin them up on demand, and DRaaS emerged in the 2010s to package this as a subscription. It brought enterprise-grade recovery capabilities (once affordable only to large firms) within reach of small businesses via MSPs.