Now is the time to think about designing high performance infrastructure with Hyper Convergence
By Murad Wagh, Director, Systems Engineering, VMware India
As the world around us is becoming increasingly digital and as most applications and services are moving to cloud promising anytime-anywhere connectivity, performance expectations of customers on their IT infrastructure are on an exponential rise. In addition, as digital touch points with customer applications go up, there is runaway data growth. Customers expect IT to keep pace with these business trends and continue to deliver flawless performance without a hiccup.
If IT takes a traditional approach in terms of adding a few more servers, a little more network bandwidth or a few more operations staff, they are in for a big surprise and it takes very little time to fall short of customer expectations. Traditional approaches tend to increase costs while increasing operational complexity. Ironically, the costs associated with increased capacity, performance and management complexity often exceed cost saving on hardware. There will be more servers to distribute across demanding applications while the network, storage and security management methods and tools continue to work in silos and hence are not agile enough to match the dynamic business requirements.
The alternative is to pause and reconsider the infrastructure basics. There are very few in the enterprise IT in Fortune 100, for example, who are unaware of server virtualization which heralded a revolution in efficient server utilization. Now, the key lies in further modernizing the data centers.
Enterprise IT needs to now extend server virtualization beyond compute to storage, network and security and manage it all through a unified management software platform. This will allow them to build infrastructure building blocks on industry-standard X86 servers with built-in storage and extend it on demand as requirements grow. In fact, entire provisioning and operation can be orchestrated through comprehensive software that embeds business requirements into a comprehensive policy. IT leaders need to take an application view of their infrastructure and design, operate and monitor it accordingly to minimize loss in translation of application demands into infrastructure specifications.
This sets the enterprise IT infrastructure on a clear path of Software-defined Data Center (SDDC). It transforms what was once static, inflexible and inefficient into something that is dynamic, agile and optimized. It helps IT take a leap from proven server virtualization to the data center of the future. In SDDC, all of IT infrastructure including compute, storage, management and network are virtualized and delivered as a service. Resources are automatically applied in response to a new demand and all of this is governed in software by the use of policies that reflect business requirements. Just as server virtualization helps in extracting the most of your computing systems, SDDC helps you make the most of all the elements, especially storage that are essential to host an application.
Traditional approach to building out infrastructure is to source the compute, network and storage components and install/network/manage manually using multiple vendor-provided tools in a BYO (Build Your Own) approach. An alternative approach is to consolidate the compatible hardware into a close physical unit but still operate most of the functions through multiple vendor tools while a few are managed in a unified manner. But, the fastest, most direct and most efficient route to SDDC is Hyper Convergence or HCI (Hyper Converged Infrastructure) where compute and storage are integrated in an industry-standard x86 server and an ecosystem of pre-certified vendors provide the necessary components as ready-to-use server nodes. The entire infrastructure uses a building block approach which brings in all the required elements together that are provisioned and operated via unified virtual management platform inside a single control software application.
Widespread adoption of server virtualization and influence of increasingly cloud-native applications demand an updated approach to IT and place it a step away from hyper convergence. This latest evolution of infrastructure management will slash operational and capital costs, enhance information technology (IT) and business agility, and improve application performance all in one stroke. HCI also minimizes risks by using existing HW, tools and skills as it scales out beyond the on-premises data center into public cloud with agility in tune with business demands.
The additional benefits of HCI are in better leveraging higher density flash storage and more powerful servers that are becoming available with new hardware innovations. In addition, increasing demands of VDI (Virtual Desktop Infrastructure), remote office/branch office deployment and management clusters also call for HCI. Simplicity and efficiency of provisioning, management and operation of mission-critical infrastructure is supported by a large ecosystem of infrastructure partners who make building blocks available in a ready-to-build server node or a totally plug and play server appliance format. Whichever implementation approach is chosen by IT, it makes it possible to manage capacity, availability/resiliency and performance through a simple software policy that goes with the application as it seamlessly moves across the infrastructure from on-premises DC, through hosted DC to various public cloud environments.
At its core, enterprise IT exists only to enable business and their critical applications. HCI is in sync with this application-centric view of the business world. You can exercise granular control over the compute resources and storage services while keeping a tight grip on performance and security – all through a single software window. HCI is not just about server virtualization or single vendor systems; it’s about using a smart set of server/storage/network building blocks that literally work the moment they are out of the box. It’s a technology that has been proven in more than 10000 IT shops of the best companies of the world.
Considering the cutting-edge business model innovations all around, CIOs need to have a finger on the business pulse in order to predict and provision equally competitive, high-performance infrastructure quickly without missing a beat on performance. They should take advantage of HCI which makes this possible by organically building on existing infrastructure and skills with ease.
