Understanding DRaaS: Part 1 - Cloud based DR versus traditional DR

Article by Nitin Mishra

In this three-part blog, we look at what is involved in designing an efficient IT disaster recovery (DR) plan and how a cloud based disaster recovery offers compelling value.

  • Part 1 - Cloud based DR versus traditional DR.
  • Part 2 - Terminology used in DR planning.
  • Part 3 - Different approaches to design a cloud based DR.

Ideally, all businesses should have an IT disaster recover (DR) plan to maintain business continuity in the event of natural or manmade disasters that disrupt their IT infrastructure.

However, traditional DR involved replicating data and applications on dedicated backup servers and storage either on-premise or at remote locations, and employing qualified IT personnel to create regular backups, test, and maintain this duplicate IT infrastructure. This was an expensive and unaffordable overhead for many companies. Worse, the whole process of restoring data from backup devices such as tapes, CD, or DVD and initializing and running applications on backup servers meant that data transfer was error prone and recovery times were slow – recovery times of several hours or few days were the norm.

Today, cloud based Disaster recovery alternatives offer a compelling alternative to traditional DR. Instead of buying and maintaining backup storage and server resources that are brought into use only in the event of a disaster, cloud computing - with its pay-per-use pricing model - allows businesses to pay for long-term data storage and only pay for servers during testing or failover. This significantly reduces cost and makes DR affordable for most enterprises.

Traditional Disaster Recovery

  • Involves a secondary physical DR site with large investments in additional data center space, connectivity and servers, and attendant operational costs in terms of power, cooling, site maintenance, and personnel.
  • A traditional DR site only operates during data replication, or in the event of a disaster. The checklist of processes involved in making such a site live inevitably involves a time lag and carries a substantial risk of loss of data and business continuity.
  • If connectivity between the primary site and the DR site fails, manual operations are required to start operations at the DR site. This can result in a significant loss of business continuity.

Cloud based Disaster Recovery

  • Cloud based Disaster Recovery services provide virtual machine snapshots of the physical or virtual servers at the primary data center. Clients only pay for storing the snapshots; storing the applications in a suspended state; and for data synchronization between the primary site and the cloud DR site. Clients pay for servers on an infrastructure-as-a-service basis only in the event of a disaster when the virtual machines (snapshots of the primary servers) need to be brought online during failover.
  • Cloud DR sites are automated and can be made live within seconds or minutes. This minimizes data loss and improves business continuity.
  • A cloud DR site can be triggered through a normal laptop or mobile device with a wireless internet connection even if the as designed connectivity fails.

Virtualized cloud platforms are well suited to providing DR. Under normal operating conditions, a cloud based DR service may only need a small share of resources to synchronize from the primary site to the cloud. The full resources required to run the application only needs to be provisioned (and paid for) in the event of an actual disaster. Automated virtualization platforms for disaster recovery ensure that additional resources can be rapidly brought online once a disaster is detected. This can dramatically reduce the recovery time after a failure, a key component in enabling business continuity.

The inherent on-demand nature of cloud computing provides the greatest cost-benefit when peak resource demands are much higher than average demand. This means that cloud platforms can provide the greatest benefit to DR services that require warm stand-by replicas. The cloud can be used to maintain the state of an application using low cost resources under ordinary operating conditions, and the more powerful and expensive full resources to run the full application need to be invoked only when disaster strikes. These additional resources can be provisioned in seconds or minutes. In contrast, in traditional DR, backup servers must be always available to address a full disaster workload – this results in significant idle capacity and cost during normal operation.

The biggest advantage of a cloud based DR service is that it moves the discussion from data centers and hardware acquisition to cloud capacity planning with all its inherent advantages:

  • Pay-as-you-go pricing can lower costs, especially for smaller enterprises
  • Automated virtual platforms minimize time-to-recovery during failover
  • Flexible cloud resources make it possible to switch from low cost to powerful servers only when needed
  • Backup data is stored in remote geographical locations
  • The cost and difficulty of working with recovered data is eliminated.