A lot of debate has arisen in the recent era of cloud computing. One of the main concerns is the choice between public, private or hybrid cloud strategy. There is no one option better than the other, but there are business drivers for each. However, generally speaking, the advantages of the hybrid cloud outweigh the private and public as it has the common benefits of both. This is especially the case for a business’s disaster recovery plan or DR plan.
However, disaster recovery plan is rarely the principal driver – although it should be. Recent studies have shown that the top reason for adopting the hybrid cloud model is principally to lower the total cost of ownership (TCO). Nonetheless, in this article, we’ll make the case that the crucial advantage of the hybrid cloud is the disaster recovery capabilities it provides. But first, we’ll briefly outline the more commonly discussed reasons for adopting a hybrid cloud solution.
Table of Contents
The common cases for a hybrid cloud strategy
As introduced, when it comes to hybrid cloud benefits, it’s usually cost that’s positioned front and centre. Equally, it’s normally cost that acts as the main reason for choosing which workloads should be moved to the public cloud and which should be kept on private servers. In terms of the kind of applications that are usually considered for the public cloud, it’s often the customer support variety as they generally have fewer security, criticality and compliance concerns. Equally, by moving customer support-like services to the public cloud, organisations can minimise the impact on the existing private infrastructure.
Interoperability is another key reason companies adopt hybrid cloud solutions. It allows for the running and management of heterogeneous applications at scale over the private and public cloud, giving the flexibility to choose the best fit for each application or workload. The competition between the different vendors for cloud services also steers the tendency toward a hybrid cloud vendor-independent model, as this addresses the risk of pricing changes (or even service providers going out of business completely, which is certainly not unheard of). With the hybrid model, you can ensure the highest data compliance, the ability to run and integrate legacy applications, store the data locally and leverage the low-cost processing resources of the public cloud.
Essentially, the hybrid cloud mixes the agility and low pricing of the public cloud alongside the highest control over data and applications provided by the private cloud, in a single workload ensuring the greatest scalability. It also ensures shared governance between a “trust me” public cloud format and the internal governance of a private cloud. In addition, it has the benefits of elastic computing to solve the complexity of setup in a private cloud and the uncertainty of how the resources are shared on the public cloud.
The hybrid cloud as the lynchpin of a disaster recovery plan
However, we’re going to argue that the key advantage of using the hybrid cloud lies in backup and disaster recovery. Why? Because it means money. Just five hours of downtime due to power loss at an operational data centre can cost millions in losses. This was the case for the world’s second-largest airline carrier Delta when it was obligated to cancel a thousand flights due to this catastrophe. Essentially, this means that you should choose a reliable backup and disaster recovery plan in the cloud – and often, a hybrid solution provides the answer.
If you’re still leaning towards a public-only solution, we’d like to remind you of the case when 20 of AWS services began failing. It affected millions of users of Netflix, Airbnb, Tinder and IMDB. Technically, this was due to the increase of error rates on Amazon’s NoSQL database DynamodB. Netflix could solve the problem as it applies a multi-region active/active replication, replicating its services between different AWS regions to allow rapid recovery from failures. Put simply, Netflix quickly directed the traffic from the impacted AWS region to another data centre in an unaffected area. As Netflix demonstrates, you need a large-scale infrastructure plan to recover from public cloud failures.
Need to hire a disaster recovery plan expert?
In contrast, here’s an example of the risks of using a solely private model. Last October, the California DMV faced a catastrophic failure because of their lack of a disaster recovery plan and proper backup. They kept the primary and the backup systems side-by-side in the same private cabinet. As a result of downtime, 188 offices around the state were so severely affected by this failure that they were not able to process any drivers’ or vehicles’ documents. In short, to ensure business continuity a hybrid cloud model should be maintained for better backup and disaster recovery plans.
Applying an effective hybrid cloud business continuity plan
We’ve mentioned some hybrid cloud examples for a proper backup and disaster recovery plan. So, what are the different scenarios for applying these strategies? In hybrid-cloud AWS, four levels of disaster recovery options or scenarios can be defined. First, there is the traditional backup and restore option to protect the on-premises data via the cloud in a disaster recovery event. Second, there is the pilot light option, which aims to replicate data and minimal running services into the AWS, ready to take over and a flare-up in a DR event. A third option is a warm standby, which replicates the data and services into AWS so it’s ready to take over. Fourth and finally is the multisite (A-Z) option, which is replicated and loaded balanced environments. This implies the active/active strategy for taking over production traffic.
The business continuity insurance increases from the first to the last scenario, till it reaches virtually uninterrupted business continuity with the last multisite scenario. The recovery point objective (RPO) can be dramatically decreased from 24 hours in the first option to less than 15 minutes in the multisite scenario. Meanwhile, the recovery time objective (RTO) almost disappears entirely in the multisite option compared to a 24 hour RTO in the backup and restore scenario. Naturally, the minimal RPO and RTO in the multisite option is very costly.
You might be wondering how RTO and RPO are calculated. There are some tools you can use, such as the Datto tool, for calculating RTO. Equally, it could be helpful to expand a little to ensure a full understanding of what RPO is. Here’s a scenario: if the last good copy of data you have is 10 hours old and your chosen RPO is 12 hours, then from this perspective you’re playing it safe in regard to business continuity. In summary, RPO refers to the point in time from which you will recover the data and the RTO refers to the point in time at which you will be up and running again.
Lately, there have been more trials carried out about applying a predictive and proactive method for calculating the RPO in complex hybrid cloud environments. Of course, the dream is to have a zero RPO and zero RTO solution, but we’re some way from this point. In essence, the RPO and RTO you chose are dependent on your business model, so you need to apply a business impact analysis to identify the acceptable RPO and RTO for you.
Azure or AWS?
However, within this context, it’s worth mentioning that Azure has some advantages over AWS. Azure includes Azure site recovery (ASR) to migrate pools of resources including Hyper-V and VMware virtualization technologies alongside the physical machine. These disaster recovery plan tools are very advantageous as when you apply a failover to Azure, you will pay only for the used resources and only when you use them. Also, Azure Hybrid Benefit enables you to use your existing server license when migrating from the on-premises environment to the Azure cloud.
Mitigate disaster, implement hybrid
To conclude, we’ll make a suggestion for a hybrid cloud model that maximises data security. It goes without saying that when applying a hybrid cloud model, disaster recovery and business continuity are critical in nature. A secure option for when a disaster strikes and production stops is to apply the disaster recovery plan on the private cloud that manages critical data. This will ensure that the RTO and RPO metrics are easily met by IT. Meanwhile, the organisation keeps its non-critical data and applications in the public cloud. Even if it has some time lag during recovery, this downtime won’t greatly affect business continuity.
Of course, cost-effectiveness is the main target for all businesses. With the hybrid model, there is no need to maintain an idle offsite production centre that may not be put to use for years. But when the business continuity and disaster recovery plan are based on the hybrid model, even if the organisation’s physical resources are destroyed, they can access the virtual machine in the public cloud and continue production. In summary, the hybrid cloud is the choice for disaster recovery and business continuity, providing a win/win approach that gives the highest security of data with the lowest cost.