In today’s digital world, change is constant – driven by the need to do things better than the competition.
This typically means achieving much more with less, as fast and inexpensively as possible.
Over the past several years, enterprises sought to achieve these goals by moving workloads to the public cloud. This move provided compute and data agility, allowing the enterprise to flex resources instantly in response to changing business needs.
While the cloud consumption model continues to offer considerable benefits, enterprises are realizing that certain applications are better hosted on-premises in a private cloud due to security, regulatory, cost, and performance issues. A hybrid cloud experience is now extremely popular, giving the enterprise the advantage of different deployment options based on specific needs of a workload. The hybrid cloud strategy offers clear benefits over traditional IT, delivering the agility and flexibility enterprises need to be competitive.
When choosing a hybrid cloud strategy, keep in mind today’s trends
For the enterprise building out their hybrid cloud infrastructure, many deployment approaches are available. When considering which is best, the enterprise should consider the following key trends
Both the edge and cloud computing are experiencing explosive growth, causing the build-out of new data centers. This growth is so rapid that traditional approaches to building and managing compute, storage, and networking infrastructure is no longer sufficient.
Data center consolidation
The need to minimize OPEX and CAPEX, while increasing the efficiency of the data center, is driving enterprises toward consolidation, virtualization, and containerization. Yet the increasing number of both data sources and data consumers is driving the growth in demand for storage capacity.
Big Data, high-performance computing (HPC), Internet of Things (IoT), and artificial intelligence/machine learning (AI/ML) workloads are placing increasing demands on data center compute, storage, and connectivity. Cloud-ready and distributed scale-out applications require more flexibility than what most data centers provide today.
Unprecedented data growth
The enterprise is also seeing unprecedented data growth and the need to collect and analyze immense amounts of data in a time-sensitive fashion.
For many enterprises, the accelerated demands of the market and competitive pressures have given rise to the requirement for continuous application delivery. The DevOps process is about tight collaboration between the business and IT operations, connecting the business directly to the enterprises’ development and delivery process.
These trends present significant challenges for the enterprise. Clearly, the data center infrastructure must change, transforming beyond traditional approaches. Instead, the enterprise needs to apply modern, software-defined implementations that provide APIs for integrating new and existing DevOps toolkits and workflows. The foundation of this type of data center should embrace an infrastructure model that can easily create and reconfigure systems dynamically – eliminating the need to reconfigure physical assets.
Composability: Can it solve the challenges inherent in today’s trends?
Composability is key in the transformation of today’s enterprise data centers, because it allows on-premises data centers to be more agile and cost-effective via a software-defined model. With composable infrastructure, enterprises can provision on-premises infrastructure just as quickly and painlessly as public cloud resources.
Composability eliminates the need to deploy and manage separate workload‑specific environments. Without having to move and reconfigure physical assets, systems are created dynamically and reconfigured through software. Enterprises can configure individual compute, storage, and fabric resources on-demand, based on the specific workload needs of the application.
Composable infrastructure also gives IT the ability to maintain physical workloads inside the exact same environment that supports virtual and container-based workloads. This ability to support multiple, mixed workloads, for both legacy and next‑generation applications, is critical to consolidating IT resources and reducing operational complexity and cost.
With composable infrastructure, enterprises integrate and manage infrastructure components through a single, unified API. As a result, the software-defined infrastructure can discover available resources, validate their configuration, and then provision just the right amount of resources required.
Composable network connectivity supports modern applications and infrastructure workloads
Because today’s enterprise infrastructure must support a wide variety of modern application and infrastructure workloads, flexibility of the network infrastructure is also vital for the modern data center. The network must be able to support a mix of workloads – from traditional applications to software-defined storage applications. Keep in mind that most public cloud solutions offer the ability to control CPU and storage resources, but they are limited because they have no control over connectivity resources.
Mixed workload traffic puts a heavy burden on the network to provide the bandwidth and response times needed to meet the service levels required by the business. Traditional network solutions were not designed to support modern application and infrastructure workload; therefore, they pose major challenges.
This is especially true for composable infrastructure because it must configure infrastructure in real time and ensure optimum performance. That means the network fabric component of the composable infrastructure must be workload aware – understanding workload performance requirements, fabric resource requirements, and intended workload behavior.
The components of the composable infrastructure leverage a common set of APIs, which enable them to meet the resource needs of a given workload. This is in stark contrast to traditional data center designs, where physical assets exist as separate, static islands with no awareness of workloads. Therefore, a critical component of the composable infrastructure is the networking component, which enables composable connectivity along with the performance of the underlying infrastructure and applications.
Transforming the data center through compute, storage, and networking composability
Many enterprise data centers are embracing composability because it provides the most efficient and cost-effective way to deliver a true cloud experience on premises. The data center network fabric is a vital component of the composable data center infrastructure. More than just connecting IT infrastructure components, composable networking fabric gives data centers the agility needed to configure and reconfigure network resources in real time without disturbing ongoing operations.
Visit HPE Composable Fabric website to read more about how composable fabric can improve your hybrid cloud network. You can also watch this short video to see how HPE Composable Rack streamlines end-to-end networking. To learn more about composability, download the analyst report, Quantifying data center inefficiency: making the case for composable infrastructure.
About the Author
Marten Terpstra is a Senior Director of Product Management at Hewlett Packard Enterprise (HPE) within HPE’s Composable solutions business unit. He is focused on developing innovative new technologies and solutions to help organization simplify, accelerate, and scale IT data center operations by delivering a cloud-like operating experience, on premises. To read more articles from Marten, visit the HPE Shifting to Software-Defined blogsite.
Featured image: ©Bigstock