Monday, Jan 23, 2017
HomeFeaturesArticlesThe Evolution of Disaster Recovery

The Evolution of Disaster Recovery

For the past decade, disaster recovery services have generally consisted of enterprises building entire backup data centers that replicate their data hundreds of miles away. Many companies today are recognizing the cost and service efficiencies of moving to multi-tenant recovery clouds, which can cut production costs in half or more and reduce the time of recovery significantly.

The Trouble with Traditional DR Models

It’s no secret that traditional disaster recovery models, where complete backup data centers are replicated, are costly. First there’s the expense of leasing colocation space and building a new colocation cage. Then comes the capital expenses to purchase new hardware, dedicated firewalls, load balancers and different security devices.  And don’t forget software licensing for virtualization platforms and additional software. Then, should an organization plan to run on the backup environment more than temporarily, they’ll need a backup for their backup. In addition to capital expenses, there are operational expenses to keep the disaster system up and running at all times. Any plan that fails to account for all of these considerations is a data redundancy solution and not a true disaster recovery program.

The costs of maintaining a backup data center can be hard to justify considering that duplicate equipment typically sits unused for most if not all of its serviceable lifetime. It isn’t used unless there is a significant outage or disaster and not merely for a short blip in service, given failing back isn’t trivial and the failover itself requires technical intervention and time.

Beyond costs, the hardware is notoriously difficult to repurpose. Nearly every administrator knows this, and most have suffered the headaches that come with recovery between dissimilar systems.

Most troubling is that these traditional back-up data centers are often slow to respond during an incident. Businesses can have their reputations ruined overnight due to service interruptions. Compounding the situation, there could be unrecoverable gaps in data during an outage if a solid continuous backup process is not in place.

Benefits of a Multi-tenant Cloud Approach

Moving to multi-tenant recovery clouds can bring cost and service efficiencies, as well as flexibility to scale and adapt as the company’s needs change.

The broader shift to cloud-based solutions gives customers more choices than they might have otherwise compared to relying solely on physical infrastructure. Given the many benefits of cloud services, it’s easy to see why recovery clouds have begun to appear in the market.

Multi-tenant recovery clouds follow an architecture approach in which a single instance of software serves multiple tenants. The software applications are partitioned to provide customized support for each tenant. Also, each tenant’s data is secure relative to other tenants. Operations such as load balancing and software management are handled by the cloud provider, simplifying system administration for the tenants. Tenants can range from dozens, or even hundreds, of companies within a single recovery cloud solution.

The primary reasons to consider a multi-tenant cloud approach include:

Cost Savings

For most companies, cost is paramount in deciding to move to a multi-tenant cloud. The cost savings of a multi-tenant recovery system come from economies of scale that are difficult to realize for a single build-out or organization.

Failing Over and Testing

Particularly when it comes to highly regulated industries that mandate testing, including financial and health care organizations, a lack of testing backup systems can leave glaring weaknesses that result in costly fines and damaging outages. The last time an organization wants to be trouble shooting its disaster recovery is during a disaster. The multi-tenant approach allows data to fail over and still perform testing, which is crucial for regulated industries. In the past, failing over was extremely laborious, but technology improvements have made things much easier.

Replication of Data

Replicating data has also come a long way. In the past, every time an organization made a patch, one had to patch both live production sites and the disaster recovery site, in effect treating the latter as a warm standby. From an operations standpoint, this meant twice the amount of work. Because the solutions of today replicate the entire machine instead of just a place to install your software, organizations only have to patch once. For smaller IT teams, this means less waste of resources. Larger teams can put their time into running disaster recovery tests, rather than ensuring the environments are in sync.

Disaster Recovery Experts on Demand

Staffing considerations are another benefit to the multi-tenant model. Depending on the nature and extent of a disaster, an organization may or may not have staff able to even log in to production facilities, let alone execute a disaster recovery plan. Those in multi-tenant environments can rely on experts who are highly experienced in recovering from a disaster and can dramatically reduce the time it takes to get back up and running. To wit, customers using cloud providers are 13% more likely to have their disaster recovery solution up and running within six hours compared to those who go it alone.

Improved Fail Back

Finally, it’s important to consider how an organization will get its data back to its primary environment after a failover. In fact, the fail back is sometimes more difficult to orchestrate than the initial failover. For a single user, traditional model, this occurs on an individual basis. Under a multi-tenant model, the process has far greater control.

Weighing Cloud Recovery Alternatives

While disaster recovery is evolving toward cloud-based solutions, some reservations remain concerning latency. This is a valid concern, given the point of disaster recovery is to minimize service disruption. However, latency doesn’t have to be a disqualifying factor in transitioning to a multi-tenant cloud model if the provider is physically nearby.

Security also is a key factor when considering a multi-tenant cloud recovery solution. As stated earlier, data remains secure between tenants despite resource sharing. Additionally, since assets are secure from tenant to tenant, multi-tenant cloud solutions actually can offer better security given most current security solutions are perimeter-based.

Another key consideration is which cloud provider to go with. It is important to check into how much data redundancy is provided in a cloud provider’s current solution. If a site is unavailable for some time, a general cloud does not have the capability to restore the data to the site and get the operation back up and running. A disaster recovery plan needs to fill in the gaps to account for how the backed-up data will actually be restored. Also, some providers operating multi-tenant disaster recovery services at extremely low prices might be running on substandard hardware, are terribly oversubscribed and may not be able to bring all customers back in the case of a failure, or may even raise prices exorbitantly in the case of a failover. With data often the lifeblood of a company, customers should be savvy when trusting their back up to the cheapest vendor.

Data recovery services have come a long way. With advances in technology, many organizations are choosing to take advantage of economies of scale and leverage multi-tenant cloud recovery solutions.