Cloud bursting is a method of application deployment. It allows an application running on on-premises infrastructure or a private cloud to extend its processing and “burst” onto public cloud platforms when the demand for computing capacity increases beyond what is available on-premises.
A cloud bursting instance can span public cloud platforms, private clouds, or on-premises computing. The model relies on the public cloud’s ability to scale on demand and offer capacity on a dynamic basis.
Cloud bursting is advantageous for organizations that have uneven or unpredictable levels of computing activity. It enables them to deliver a consistent user experience, regardless of load.
Examples include:
Rendering of images or video, such as in computer-generated imagery (CGI). This process creates large spikes in computing after periods of relatively low computing activity.
eCommerce transactions that grow quickly during a sale, exceeding the capacity of primary computing instances set to handle shopping volume.
Data analytics or artificial intelligence (AI) computing cycles that grow substantially during training or pattern recognition sessions.
There is more than one way to implement cloud bursting. The process can be manual, wherein teams provision and de-provision cloud overflow instances by hand. This approach is generally for seasonal use cases when an organization can predict the need for cloud bursting in advance. Alternatively, with automated cloud bursting, a specialized solution establishes parameters for bursting, such as when compute reaches 50% of capacity and then automatically diverts the computing load to the cloud when the parameters are met.
A third approach is to distribute load balancing between on-premises and cloud instances. In all three cases, the infrastructure supporting cloud bursting may vary based on a hybrid or multicloud approach. For instance, bursting can be from cloud to cloud in a multicloud architecture.
Cloud management platforms like Nutanix can set up and orchestrate cloud bursting. For example, Nutanix NC2 enables users to rapidly burst capacity from their datacenter to the Amazon Web Services (AWS) cloud platform as needed.
Let’s see how cloud bursting can work for you. Contact us today.
A hybrid cloud, which comprises connected instances of on-premises and cloud compute, is the predominant approach to cloud bursting. In cases when on-premises compute exceeds its processing capacity, the workload bursts to the cloud.
Hybrid cloud is not the only way to execute cloud bursting. A multicloud architecture, which spans multiple platforms like AWS, Microsoft Azure, and Google Cloud Platform (GCP), will also work. In that scenario, a workload bursts from one cloud platform to another. Why would someone want this? The answer is that robust infrastructure, which can be realized by leveraging more than one cloud platform, helps ensure positive outcomes for cloud bursting. It enhances business continuity through greater resiliency. If one cloud platform fails, another takes up the workload. Additionally, using multicloud for cloud bursting helps avoid vendor lock-in in the cloud.
Cloud bursting offers many benefits to organizations that deploy it. Key examples include:
Planning for scalability - Cloud bursting can be a forcing function that pushes application owners and other stakeholders to assess resource demand and engage in capacity planning. This process helps prevent outages and other disruptions due to overload. Planning for cloud bursting also requires relevant teams to take application scalability into account during the design and deployment stages of the application life cycle. This way, they can rely on auto-scaling services and tools in the process.
Effective cost management - While cloud bursting comes with a price tag, the method has the potential to save money by helping avoid over-investment in on-premises infrastructure. For example, procuring all the capacity needed for the peak season and letting it remain unused for the rest of the year will result in wasted capital expense (CapEx). A thorough cost analysis is required, however, including assessments of pay-as-you-go models versus fixed commitments to cloud capacity. Monitoring and controlling costs should also factor into the return on investment (ROI) analysis for cloud bursting.
Ensuring disaster recovery - Cloud bursting can support a disaster recovery (DR) strategy. This might involve automating failover from on-premises compute instances to the cloud in the event of an outage. Alternatively, by continuously replicating data to the cloud, i.e., “mirroring,” cloud bursting enables high availability and DR.
Managing peak load challenges - A well-planned and executed cloud bursting implementation enables stakeholders to predict and prepare for peak computing times, such as seasonal sales. Cloud bursting also makes it possible to implement real-time traffic management that can handle unexpected spikes in demand and preserve application response times for a good user experience.
Optimization for application performance - A cloud bursting solution may include functions for application performance monitoring and optimization. By monitoring application performance levels and triggering cloud bursting when necessary, an organization can optimize application performance.
Strengthening data security and compliance - Cloud bursting can be beneficial for data security and compliance, assuming proper policies and controls are in force. Any cloud bursting deployment, however, should be subject to careful security and compliance review, especially in multicloud architectures.
Ready to experience these benefits? Contact us today.
Cloud bursting can be challenging. For instance:
Nutanix solutions, including Nutanix Database Service (NDB), Nutanix Unified Storage (NUS), and Nutanix Cloud Manager (NCM), offer features and functionality that help application owners, security teams, and cloud managers overcome these challenges.