In our previous posts, we covered the disciplines of Business Continuity and Disaster Recovery from a theoretical and academical perspective. Today we will cover Cloud-based Disaster Recovery, and why it makes sense to use such a service.
Of houses and data centers
I like to make the parallel between primary data center sites and secondary data center sites (those being often used for DR purposes) as the equivalent of owning your own residential home (where you live on a daily basis) and owning a holiday home as well. It’s perfectly fine to own both, especially if you have enough wealth to take care of these, but as a matter of fact you are going to spend most of your time in your residential home, while your holiday home will eventually have the honour of your visit a few times a year. Maintaining two properties and whatever is tied to them comes at a cost increase, whether it is utilities, mortgage, maintenance etc, and especially your holiday home stands there, unused except for a big holiday a year or a couple weekends if you’re lucky. You can of course decide to rent your holiday property but you cannot exactly predict what will happen, how visitors will take care of your property and if you’ll incur extra costs. Plus, if you don’t live in a large family, you will likely be the only person responsible for taking care of both.
The challenge with secondary / Disaster Recovery data centers
As you can see, it’s very easy with this example to draw the parallel with data centers. With one twist, however, while your holiday home can accommodate very simple and spartan furniture, your secondary data center cannot: it requires physical infrastructure to be present, configured and ready to kick into action in case a disaster happens at your primary data center. That means compute and storage to run the workloads, network infrastructure in place, WAN connectivity between your sites and also anything relevant to your backup / DR infrastructure (backup servers and eventually backup targets). Finally, maintaining a secondary data center means not only owning or renting the premises, but also in environmental costs: energy bill and power infrastructure (including uninterruptible power supplies and eventual power generators), cooling & climatisation, monitoring & security.
The strategy of operating one’s own primary and secondary data centers (especially when those are used for Disaster Recovery purposes) has been and is still a widespread concept. After all, it’s only a few years that using cloud has become mainstream, and many companies have been relying on their pre-existing processes and the confidence that « we’ve always done it that way » cannot be wrong. Add to this the reassuring but illusory feeling that owning and operating infrastructure is better, and you get to the current situation.
Cloud Disaster Recovery: all the benefits of in-house DR while saving costs
Cloud Disaster Recovery solutions are to be considered as a true replacement of your secondary locality / Disaster Recovery data center. The entire infrastructure stack up to the DR solution is managed by the provider and allows the customer to focus on what really matters: technically implementing an effective Disaster Recovery strategy, without having to worry about managing infrastructure and costs associated with it, or maintaining special dedicated WAN lines – all you need is an internet connection to the cloud DR provider (and of course a VPN to secure communication). Furthermore, a vast majority of cloud DR providers implement a « pay-as-you-go » consumption model, which further helps to keep costs under control.
Financial aspects aside, customers leveraging a cloud-based Disaster Recovery solution should rightfully expect it to operate the solution in a nimble way, and thus at least as well as if this was their own infrastructure.
The following characteristics should be supported:
- Hypervisor agnostic support – Industry standard VMware and Hyper-V virtual workloads should be supported
- Support for physical workloads – backup and replication of physical workloads through a P2V process -and spinning them up in the cloud- is not only necessary but also a critical differentiator between services
- Instant recovery – the goal of DR is to spin up the protected environment in the cloud instantly after a failure/disaster has happened
- Automated IP re-addressing – to automate the DR process as much as possible
- Multiple geographical locations – to support companies operating globally
- Encryption of replicated data – this is crucial to ensure that your data is inviolable even in the event of a data breach
- Dedicated management portal or GUI – you should be able to operate your environment in an easy way, without having to know the intricacies of the underlying infrastructure
In a time where IT budgets stagnate (a fact that has been consistently true for several years now) and where IT departments are asked to do more with less, reconsidering existing DR strategies can be a powerful factor in cost reduction and increasing operational efficiency.
Leveraging Cloud-based Disaster Recovery solutions that integrate with existing infrastructures allows for drastic cost reduction by eliminating the need to operate secondary data centers & additional infrastructure with all the associated costs, including infrastructure lifecycle refreshes every five years.
Getting back to the analogy we presented in the beginning of this article, leveraging cloud Disaster Recovery should be seen as consuming a service, and should be as easy as renting an hotel room for a holiday. You don’t have do worry about the hotel operations when you take a holiday, and you shouldn’t have to. The same should apply to using a cloud-based Disaster Recovery solution.