By Shiji Sujai, IOD Expert
I first came across a full-fledged DR Suite almost 10 years ago when the organization I was working for decided to implement a real-time DR solution for our main data centers. I had my apprehensions: why did we need such a costly and complex solution when the backups were all working fine?
Blame it in on my rookie thought process of the time. Backups serve the cause of business continuity, but at two different levels depending on the RPOs and RTOs to be achieved. This is something I realized only much later, though, when I saw this beautiful technology working its magic through real time replications and failovers between data centers miles apart.
Years passed, and organizations are now shifting focus from owned or co-located data centers to cloud. Why make a huge capital investment in all the equipment, power, cooling, and rent when the same capabilities are offered at a fraction of a cost by public cloud service providers on a pay-as-you-go basis? (This question is specifically relevant for data centers maintained only for DR purposes.) Before we discuss the whys and hows of moving your organization’s DR to the cloud, let’s review the terms and practices used related to business continuity.
RPO and RTO
Recovery point objective (RPO) is focused on the data that your organization can afford to lose should a disaster strike and is dependent on the interval between your backups. Therefore, organizations often design their backup frequency according to the RPO of an application. For example, if the RPO of an application is 15 minutes, you should schedule a backup of the application to happen every 15 minutes. Otherwise, there could be serious consequences to your agreements with customers when the service is back online after a disaster. If the application in question is, for instance, used in a financial or banking industry vector, your company could face litigation.
Recovery time objective (RTO) is focused on the maximum tolerable downtime for your applications in the event of a disaster. This is in turn linked to the availability SLAs with your customer, and failure to recover within the defined RTO could result in financial penalties. It also implies that having a backup solution matching your RPO is not enough, unless you have a robust solution that can recover the data and make your solutions online within the stipulated RTO.
Isn’t Data Backup Good Enough?
Often data or application backups are good enough, when you still have a platform (a.k.a. data center) to host those applications and recover them. However, data backups cannot hold down the fort in the event of a catastrophe where the data center is inaccessible, let alone your backup tapes or disks that hold your data.
Many organizations have an offsite backup storage location. How fast you can recover using an offsite backup depends on how often the offsite backup is taken and transported offsite. What if offsite is also within the vicinity of the natural calamity that impacted your data center?
DR solutions are designed at a more advanced level than traditional backup solutions. Data is replicated to a different location, most often in real time. The control plane to trigger the failover should ideally be independent of the source location. This helps in initiating failover when there is no connectivity to the source which is most often the case in the event of a disaster. The only caveat of traditional DR solutions is the requirement to build and maintain a data center in standby just for Disaster Recovery. It is an idle investment, but something that organizations cannot eliminate if they want to meet the defined SLAs with a customer. Over the past few years, cloud service providers have been offering a cost-effective alternative where organizations can leverage compute, storage, and network capabilities in the cloud on a pay-as-you-go basis to meet business continuity requirements.
Why Should You Move Your DR to Cloud?
DR solutions are complex. I remember that it took weeks for us to put a DR solution into place for our main data centers and additional weeks to test and ensure that it was working fine. Testing the solution was something we could do only over a weekend, as we couldn’t bring down production systems during a week day to play around with DR. A lot of assessment and planning went into the project, as we had to make sure we did a failover and again failback to the original data center before business resumed on a Monday.
When I think about this process now, it seems prehistoric as we have smarter and more efficient DR solutions available, directly integrated with cloud and capable of all this in a matter of hours. Physical media management is always the difficult part in both backup and DR solutions as you need to make sure they remain usable all the time. Organizations commonly use disks or tapes for storing the data offsite. Often it could be a single copy of data which, if not retrievable, becomes a single point of failure in DR scenarios. Cloud-based DR solutions or DRaaS service providers leverage cloud storages that are backed by almost 100% SLA and redundancy already built-in. For example, Azure cloud storage always has three copies of data replicated and available within a data center.
The Many Perks of Cloud-based DR
Ease of configuration is a major plus point for cloud-based DR solutions. While using hosted DR solutions, you can plug in your environments by configuring few additional components. Enabling DR protection using an Azure-native DR solution, Azure Site Recovery (ASR), for your Hyper-V VMs is as easy as installing an agent in the host server. Most cloud-based DR solutions offer secure transfer and encrypted storage of data in cloud thereby addressing security and compliance concerns of organizations while storing sensitive data in the cloud. Last but not least, these solutions offer a non-disruptive way of testing your DR strategy without affecting your production environment. Leading cloud-based DR service providers like Carbonite double take, Veeam and Azure Site Recovery (ASR) offer this capability.
How to Move Your DR to Cloud
Azure Site Recovery is a first-party cloud-based DRaaS solution from Microsoft Azure that can easily integrate with heterogeneous environments and enable a fully functional DR solution with no capital investments. ASR can be used for Hyper-V and VMware based virtual environments as well as physical servers with support for Windows and many flavors of Linux operating systems. Backed by trusted Microsoft technologies, minimal configuration is required to integrate services with ASR and Azure cloud acts as your secondary data center in DR scenarios. There are options for failover, failback and test failovers so that you can always be double sure of your DR strategy being effective in the event of a real calamity. As explained earlier, the data is stored in Azure storage which is secure and resilient by nature.
ASR is the service of choice for most enterprise customers as it can cater to multiple use cases and scenarios, but there are other third-party niche players who provides cloud-based DR solutions often catering to specific virtualization platforms. For instance, VMware offers VMware site recovery for protecting on-premise workloads to VMware cloud in AWS. Veeam on other hand supports replicating Hyper-V and VMware workloads to AWS, Microsoft Azure, and IBM cloud.
Better safe than sorry. These words always hold true when it comes to DR strategy of any kind. All organizations large or small should have one in place in preparation for inevitable yet unexpected downtimes or disasters. DR solutions have clearly gone through a makeover since my IT admin days: if a DR solution is not built to leverage the vast capability and scalability of the cloud it’s not good enough. Organizations should adopt a cloud-based DR solution provider to future-proof their DR strategy.