Five key questions about disaster recovery as a service

What is DRaaS?

The growth in cloud computing – and especially cloud storage services such as Amazon’s S3 and Microsoft Azure’s Blob – has changed the face of the backup and recovery market. A growing number of suppliers offer disaster recovery (DR) services either entirely in the cloud or combined with on-premise technologies.

The cloud gives organisations an easier way to create off-site copies of their data, or even replicate entire production environments. It does away with the need to acquire datacentre or colocation space and to build a secondary failover system.

In addition, with more workloads now in the cloud, it makes sense to failover from one cloud environment to another.

In both scenarios, DRaaS promises to bring an organisation’s systems back online quickly, without the need to source new hardware and facilities, and with minimal disruption to workflows.

DRaaS, however, is different to simple cloud-based backup services. Due to the complexity of disaster recovery, there is far more emphasis on services tailored to the customer’s production environment and business processes.

DRaaS or BaaS?

Backup technology suppliers, storage suppliers and the large public cloud providers offer a range of backup and disaster recovery services.

Backup is an obvious use case in the cloud because it has the ability to scale and allows firms to move to a pay-as-you-go consumption model for data storage.

In addition, cloud backups fulfil the requirement for business continuity to keep at least one copy of data off-site, and can provide an “air gap” to protect data from attacks such as ransomware.

But, backup as a service (BaaS) and DRaaS are different services. BaaS is the better-established, with the focus on keeping one or more copies of data secure in case of data loss. Suppliers now offer sophisticated levels of tiering for cost and performance, to allow economical, long-term data storage. This is especially attractive in regulated industries, where this also meets compliance needs.

Recovery is usually to the firm’s on-premise hardware, but this process can be slow. Even recovery to cloud infrastructure means starting from scratch – spinning up new cloud instances and virtual machines (VMs), installing applications and restoring data.

DRaaS overcomes this by creating a copy of production data and systems in the cloud. These replicated systems are similar to the hot standby hardware used in industries that need high levels of data availability and business continuity – but they bring the cloud’s economies of scale and remove the upfront hardware costs of conventional redundant systems.

DRaaS also has the advantage of running as a service. There should be little need for intervention by the IT department once it is up and running. Suppliers also offer additional services, such as DR testing and consultancy. Of course, these services come at a cost.

Who uses DRaaS?

Almost any organisation can use DRaaS because it requires little in the way of hardware or up-front investment.

However, its use is most common in organisations that want to minimise downtime, but cannot justify investment in redundant hardware, either on-premise or in a datacentre or colocation facility. This is likely to involve a trade-off between performance and recovery times, and cost. DRaaS that runs in the public cloud will be slower than dedicated systems, but it will still be faster to recover from than basic cloud-based backup or BaaS.

Another application for DRaaS is where conventional DR systems are less practical. This includes branch and remote offices that may have lower bandwidth connections and little in the way of on-site IT support.

There is also a trend towards use of DRaaS to provide resilience for cloud-based infrastructure. Such cloud-to-cloud disaster recovery can range from replicating entire cloud production environments or specific VMs to a secondary cloud location, to providing additional redundancy and continuity for SaaS applications and even Microsoft 365.

Some firms create their own cloud-to-cloud disaster recovery systems, but the trend is to buy this in as a complete service.

Costs of DRaaS

Suppliers use a range of charging models for DRaaS and that can make cost comparisons difficult.

Most suppliers charge per gigabyte of data protected, per month, which is a similar pricing model to BaaS. However, some charge per server protected, or per socket for virtual appliances. For desktop or end user device protection, fees can be based on the volume of data protected per device, per month. Some will also charge a fee for any data restored.

One way DRaaS suppliers can save firms money is by pooling the volume of data protected rather than forcing users to pay a fixed sum per device.

However, as Freeform Dynamics analyst Tony Lock points out, more suppliers now offer fixed monthly fees for DRaaS. This can work out cheaper than usage-based pricing, especially when a DR plan is invoked.

For organisations that build their own DRaaS, the main costs will be cloud storage, licences for software used to create and manage the replicated systems, and staff time to build and maintain the system. CIOs also need to watch out for variable costs across storage tiers, and for data egress charges in the event of a recovery situation.

Who offers DRaaS?

A wide range of suppliers offer DRaaS. These include AWS and Microsoft Azure. Backup providers such as Acronis, Commvault, Unitrends and Veritas have DRaaS products, as do storage suppliers such as Dell and NetApp.

Other notable suppliers include Carbonite, Cohesity, Druva, N-able and Veeam. Specialist service providers such as 11:11 and Databarracks are also active in the DRaaS space.