This content is part of the Essential Guide: A DRaaS market guide: Advice on the thriving technology

Cloud recovery service providers: Areas of concern

Discover why location, networking, failback and SLAs are important in picking a cloud disaster recovery service. Then find out what it takes to implement it.

There are multiple service offerings in the market today for implementing cloud disaster recovery. These range from pure replication tools that get data or a virtual machine image into the cloud, to fully managed DR as a service capabilities. When looking at cloud recovery service options, you need to consider the following:

  • Location. Where is the cloud recovery service located? The further away the DR service is, the greater the latency to get data there and back -- which can impact replication performance. There is also a compliance consideration. Will the data be located in a country or region that clashes with the regulatory requirements for your data? You also need to think about whether the cloud DR vendor has itself implemented disaster recovery, in the event that there's a need to run for a number of months in DR.
  • Networking. Will the network have sufficient bandwidth for your needs, and how will you be charged? How will you integrate your network with that of the cloud recovery service provider? Are there security tools and processes to ensure your network is isolated from other customers' networks?
  • Failover/failback. Getting applications into the cloud requires some kind of seeding process. This may be achieved over the network or may require the use of physical backups that can be shipped and uploaded by the cloud provider. Look at how quickly applications can be on-boarded to the cloud-based disaster recovery service. Failback is tied to failover/seeding. At some stage, applications need to be moved back to their production location. Any data updates and operating system patching need to come back with the data, and this can take some time to achieve. Part of the DR plan may be to only ship application data and run a duplicate application in the cloud to meet service-level agreements.
  • SLAs. In addition to recovery point objectives, recovery time objectives and service-level objectives, cloud DR vendors should offer SLAs. These will determine parameters like network response times, how quickly DR can be invoked and the process for doing so. There should also be SLAs in place for cloud hardware availability -- moving to a DR platform that offers worse availability than the primary site won't be popular.

Read part 1

This is the second part of a two-part series. The first part explains why you need to build your cloud DR plan before selecting a cloud recovery service provider.


Once the background planning has been done and a cloud recovery service chosen, allocations need to be moved to the new DR process. This may require some additional work to implement, such as:

  • Virtualization of bare-metal servers. Some applications may need to be migrated to virtual machines, with the move to cloud DR providing the opportunity to justify the work.
  • Changes to existing backup processes. The specific backup tools in use may change, or be augmented with cloud DR if two backup methods are to be used. Operational processes will require review and amendment to be in line with the new backup system.
  • Changes to infrastructure. Some additional services may need to be built, especially around DNS/DHCP/AD. Networking changes may be required to allow traffic to flow between the on-site backup systems and cloud backup provider.
  • Testing. The DR process should be tested regularly. Cloud-based DR provides an easier way to test disaster recovery as the cloud-based image can be isolated more easily on the network, without impacting the production service. Many cloud recovery service providers have this facility as part of their offering.

Finally, the DR plan and process should be reviewed and tested regularly to ensure that it still delivers a fit-for-purpose product. All DR products should be seen as an evolving system that changes to match the requirements of the applications and, ultimately, the needs of the business.

Next Steps

Cloud recovery service providers take advantage of removable media

Important to test your cloud DR service

Plan to fail back after cloud recovery situations

Get the most out of your cloud recovery service

Common cloud DR misunderstandings

Test your knowledge with this cloud DR quiz

Dig Deeper on Disaster recovery planning - management