Adopting cloud in your back-up strategy

In order to enable organisations to compete effectively in this demanding marketplace, CIOs are looking for increasing levels of performance from their infrastructure. At the top of their lists are ways to make infrastructure more dynamic, resilient and to take advantage of virtualisation, which itself is changing the traditional computing model.

Cloud computing offers a way to achieve many of these goals and is increasingly becoming part of corporate strategies.

Cloud computing defined

As an aside, it's advisable to be aware that there are many definitions of cloud computing - so many that the US Institute of Standards and Technology (NIST) has issued a formal definition: "Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (networks, servers, storage, applications and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction."

As a model for consuming and delivering infrastructure, the cloud enables self-service, different sourcing options and significant economies of scale. It's also clear that organisations will use a combination of private and public clouds to achieve their goals.

When it comes to back-up, cloud has many attractions for CIOs, who face the challenge of increasing volumes of data and static IT budgets, allied to growing requirements to keep vital data and the systems it runs on available. In other words, traditional back-up, which protects the data, is no longer sufficient - complete IT resilience is essential because without it, very few organisations can survive for extended periods.

Cloud-based back-up also has the advantage of giving peace of mind because it occurs off site and reduces the need to do tape back-ups.

Caveat emptor

The old Latin tag, "buyer beware", is particularly relevant here because many services that are marketed as disaster recovery are actually nothing more than offsite data copies. That is, they replicate data but not the systems on which the data, applications, configurations and operating systems reside. Data replication is no longer acceptable for most businesses. In today's global economy it is expected from organisations to be 100 percent available at all times.

By contrast, a true disaster recovery mirrors the entire system as well as providing the infrastructure on which to bring up the systems. A cloud-based solution will leverage the syndication of such systems that the service provider can offer due to multiple customers that share the same infrastructure.

A complete disaster-recovery solution as described above is costly. It requires large volumes of data to be copied across the public Internet or via a private WAN link. A lot a bandwidth is required to ensure the Recovery Point Objective (RPO) remains at acceptable levels and the solution is reliable.

It is also important to optimise data replication by using compression devices, selective replication and incremental copies.

Prioritising data and systems for replication is, thus, key - and to do so one must understand the organisation's risk profile and identify critical data and systems. To prioritise effectively, it is critical to perform a risk assessment, as well as business impact and infrastructure impact analyses.

Once this homework has been done and you have a full understanding of the hierarchy of data and systems in relation to business criticality, then the business recovery plan can be created. And as part of this business recovery plan, cloud's flexibility and ability to reduce costs will likely be an attractive solution.