What if you could reap the benefits of cloud computing without going through a big cloud migration first? It is a well-known fact that cloud computing can give your business near-unlimited resource flexibility, top notch pay-as-you go managed services, and access to new advanced capabilities. Yet, a lot of organizations are still grounded to their own servers and data centers. Either by business choice, security concerns, or lack of opportunity to execute a cloud migration, an organization can still quickly start to leverage the benefits of cloud computing by pivoting to hybrid cloud management with cloud bursting.
In this post we’ll take a closer look at bursting to the cloud and how it can benefit your organization with the help of NetApp Cloud Volumes ONTAP.
Cloud bursting is a hybrid architecture model designed to give applications located in physical data centers on-demand access to the cloud resources, no migration needed. This enables applications and data that exist in the corporate on-premises infrastructure to seamlessly leverage cloud computing.
Cloud bursting can come in handy, for example, when a sudden traffic spike demands additional resource capacity on the fly. Cloud bursting also provides access to other cloud-based services for AI, analytics, and more.
The term cloud bursting was coined in 2008 by Jeff Barr, AWS Chief Evangelist, to describe an architectural pattern he observed in organizations. At that time, the cloud was a fairly new concept and its adoption divided opinions between enthusiasts and skeptics. When these two groups co-existed in the same organization, that often led to a compromise resulting in a cloud bursting architecture.
The AWS cloud bursting concept has become quite popular for companies wanting to kickstart their cloud journey. Over time, the term ended up being adopted industry-wide and by other vendors. The cloud bursting architecture pattern combines the existing baseline business critical infrastructure with a new cloud-based infrastructure capable of providing extra resource capacity on-demand.
A typical cloud bursting use case is for applications that can benefit from elasticity (e.g. e-commerce / sales portal) and want to rapidly answer an unexpected customer demands (e.g. black Friday sales), directing the overflow traffic to the public cloud, and thus avoiding service disruptions and maintaining extra capacity on-prem. Another typical case for cloud bursting is in organizations looking to use their application data to support new strategic use cases, such as analytics, AI, or machine learning. With a cloud bursting architecture, the application data can remain on-premises and the cloud managed data services, which offer out-of-the-box advanced functionalities and the pay-as-you-go model, can use the data on-demand as needed.
There are several immediate benefits to start cloud bursting in cloud computing. Instead of the lengthy process of a full cloud migration, a hybrid cloud bursting architecture can be implemented in a short time and without disrupting the existing systems and application business logic. This architecture type opens the door to a whole new range of cloud managed services that can be used to extend your existing on-premises functionality.
On the other hand, a cloud bursting architecture comes with some limitations.
One issue is that in a hybrid architecture, the split between different geographical locations can translate into performance lag due to the inherent latency and data storage locality.
Another potential limitation of the cloud bursting model is that some cloud services are specifically designed to be used within the cloud provider (e.g., assigning IAM roles to resources)—trying to make them work with your existing on-prem systems may pose difficulties an all-cloud architecture would avoid.
Regardless, cloud bursting remains a good option for organizations looking to expand to the cloud with the minimal amount of modifications to the existing process, automation, and tooling.
While there are native cloud services designed to manage hybrid cloud data management, such AWS Storage Gateway, these tools are more focused on enabling cloud adoption by cutting storage costs and avoiding re-architecting applications—they are not suitable for the cloud bursting model’s on-the-fly consumption of resources.
NetApp Cloud Volumes ONTAP is an enterprise-grade data management platform that enhances the cloud provider's storage capabilities. By using Cloud Volumes ONTAP in AWS, Azure, or Google Cloud Platform, organizations can benefit from features such as high availability, data protection, cloning, and storage efficiency in their cloud environment. For workloads that require or benefit from reduced latency, Cloud Volumes ONTAP can be combined with Intelligent NVMe caching capabilities, increasing the performance of your cloud storage resources.
For organizations interested in cloud bursting, Cloud Volumes ONTAP utilizes FlexCache®. FlexCache is NetApp caching technology that enables sparse, writable replicas of Cloud Volumes ONTAP volumes, bringing data closer to on-premises systems with a fast throughput and minimal footprint. This feature is specifically designed to solve the data silo, performance, and management issues that often affect hybrid architectures. As such, FlexCache is a huge enabler for a cloud bursting architecture, since it significantly increases performance and simplifies the overall setup. This provides significant added values when it comes to performance, flexibility, and your overall cloud capabilities.
A FlexCache volume is mounted on-premises and paired with a Cloud Volumes ONTAP volume. This provides on-premises applications the ability to access data in ONTAP volumes while taking full advantage of capabilities such as increased performance, caching, data protection and storage management.
How does this technology work in practice? Let’s take a look at some customer case studies.
The first organization that we’ll look at is a leader in the semiconductor industry that was facing a severe computing capacity problem. In this industry, new projects are often data-heavy—ensuring extra storage space and computing resources are available is key to success. In semiconductor projects, there are several data challenges involved, with multiple read-intensive workflows that require data collaboration and large datasets to be easily available. Since this organization had a lot of on-premises environments and needed easy access to cloud data, that made a perfect case of a cloud bursting architecture.
However, that option still didn’t tackle the challenges of data copies, data syncs, and performance. This is where NetApp Cloud Volumes ONTAP came in and made the difference. Combined with FlexCache and intelligent NVMe caching, they were able to do cloud bursting without compromising performance, and depending on the use case, even outperform the AWS native services, achieving a read-access as high as 90-95% of what it was on-premises. With the company moving 100-150 petabytes of storage on Cloud Volumes ONTAP and using FlexCache on-premises, they no longer need to mirror entire datasets between cloud and on-premises, reducing up to 90% the data per workload. The Cloud Volumes ONTAP volumes were easy to set up and automate, giving them maximum flexibility to scale as needed.
Another organization that decided to cloud burst with Cloud Volumes ONTAP did so with a very different motivation. This financial company had been transitioning to cloud over several years and was facing an urgent need. One of their business critical on-prem workloads that performs data analysis on investments had reached computing capacity. They were looking for ways to leverage cloud computing resources, without having to do big modifications on processes, automation, and skill sets, all while supporting their cloud strategy.
Once again, FlexCache and Cloud Volumes ONTAP was a critical key piece in the solution. After extensive performance testing, it was adopted in the organization and ended up as part of their company’s baseline infrastructure-as-a-code reference templates. In addition to extra computing power for that critical workload, the organization is now looking to roll this same approach across several other analytics workloads and leverage cloud managed services.
Choosing a cloud bursting architecture is not a silver bullet for cloud adoption. The limitations and drawbacks need to be carefully weighed in. However, it is an important architectural scenario that can and should be used to accelerate your adoption, gain flexible on-demand extra capacity, and leverage cloud managed services for use cases such as data analytics and data science.
Being able to start leveraging cloud infrastructure in a short period of time with cloud bursting—even in complex legacy applications—is a great value proposition. But the performance issues that come with this model can bring complications that can hinder business operations. NetApp FlexCache with Cloud Volumes ONTAP can give an extra performance boost and advanced data management capabilities to overcome the cloud bursting challenges and maximize your IT investment and technology capabilities.