When value added resellers approach the task of implementing an integrated data storage project, chances are they’ll want to discuss with their customers the benefits and drawbacks of implementing either a converged or hyper-converged infrastructure.
Server and storage platforms and the technology that supports these systems are constantly changing. Big data continues to grow and storage demands are increasing. Added to this, advancements in software have enhanced data center computing functionality, while cloud computing continues to be an attractive option for many businesses. In this environment customers are weighing their options and looking for the best IT infrastructure that supports their growing workflow and data needs.
These shifts in technology are causing VARs and their customers to navigate the plethora of hardware and software options available for their projects. A conversation about whether to choose a converged or hyper-converged infrastructure can help customers figure out what their requirements are and what technology infrastructure can best help them meet their storage, networking and computing needs.
In a converged infrastructure model that is hardware centric, servers, data storage devices and networking equipment are bundled together. This packaged solution, which is pre-configured and shipped to the customer, also includes software that automates the IT infrastructure management, and provides support for workflow and orchestration. Also of note is that in today’s converged environment, a vendor and the vendor’s partners provide bundles of hardware and software in a single chassis.
A converged infrastructure is designed to consolidate systems and centrally manage IT resources, while raising resource utilization rates. For customers that want to minimize compatibility issues and simplify the management of their servers, storage hardware and networks, a converged infrastructure is an approach to data center management that could be a good fit for them, especially if they don’t think their data and workflow needs will change significantly over time. Furthermore, customers can benefit from cost efficiencies that come with less cable connections, reduced power, less floor space and fewer IT managers to operate a converged infrastructure.
Still, a converged infrastructure model has its drawbacks, and customers will have to decide whether they want a pre-configured system. A pre-defined configuration in which each component is selected and configured by the vendor limits an IT manager’s ability to select a router or a storage array that is tailored to suit their needs.
Another disadvantage is that IT managers must maintain support of the system by updating patches in the pre-configured system, but patches are updated on the vendor's timetable and not necessarily when the user requires the patch to fix a computer program or its supporting data to improve the usability or performance of the system.
If customers perceive that their storage and server platforms require a more robust IT infrastructure to manage more data and workload then they may prefer a hyper-converged infrastructure.
A hyper-converged infrastructure takes the converged infrastructure model to the next level. In a hyper-converged system, compute, storage, networking and virtualization as well as other data related technologies are tightly integrated. In this model, which is software centric, physical servers have a virtualization hypervisor that operates each virtual machine created on that server.
By adding nodes - the hyper-converged architecture has the storage controller function running as a service on each node in the cluster - customers can realize IT scalability, which inherently implies that a business’ underlying technology infrastructure can accommodate business expansion.
Furthermore, because of the tight integration of software components that hyper-converged systems have, customers can expand beyond storage, compute, networking and virtualization, enabling them to also include data management technologies such a wide area network (WAN) optimization, compression and data deduplication.
A hyper-converged infrastructure does have shortcomings. An IT manager that wants to make small upgrades will find this to be a challenge. Looking for software that supports configurations in a hyper-converged architecture is a pain point. Additionally, an IT manager that finds a cluster running low on storage but not on compute, may have to add another appliance to upgrade compute capability.
As customers think ahead about how they’ll reengineer their IT infrastructure, they’ll have to make predictions about how their business will grow. That discussion should include key elements of what a converged and hyper-converged infrastructure can do for a customer’s business.
While both approaches share common goals such as reducing operating costs, increasing efficiency, usability and agility in storage and server platforms, there are significant variations in these two approaches. VARs can add value to their customer engagements by developing a strategy that identifies an organization’s business goals and matches those goals to the potential benefits (and drawbacks) of implementing a converged or hyper-converged infrastructure model.