Converged infrastructure is one of the hottest new data center concepts today, enabling organizations to manage resources, control users and deliver rich content in new and more efficient ways—and helping to create real competitive advantages for the business.
But what does that really mean? How does one converged infrastructure differ from another? What are the business and data center considerations around the technology?
Before jumping into the converged infrastructure pool, there are three important factors that organizations need to understand.
1. Not all converged infrastructure systems are created equal.
Organizations need to do a great deal of research when evaluating converged infrastructure as there are a number of different elements that differentiate one converged platform from another. The most important thing is to have a clear understanding of your organization’s goals for converged infrastructure. Only then can you determine which features will—and won’t—make a significant difference.
When working with converged systems and virtualization, for example, you need to know which hypervisors are supported. Some converged platforms support only one type, while others can work with multiple types. Some will even allow virtual machine translation and storage repository movement among heterogeneous hypervisor architectures.
Companies also need to consider the features that converged systems providers focus on. Some hang their hats on better deduplication rates and I/O coalescing, while others tout their capabilities in the areas of data encryption and open API support. Compression ratios, capacity delivery and cloud support are other areas on which providers commonly focus.
Hardware support is another important differentiating factor. In some cases, a converged system will allow you to “pre-build” a stack of hardware and utilize the disks and resources therein. In other cases you can go completely “white box” and aggregate existing storage systems. The great part about converged architecture is that it will integrate with your hypervisor, self-provision resources, and enable organizations to get up and running quickly. In any case, be sure that your version of the hypervisor will work with your converged infrastructure ecosystem, and that your hardware can appropriately support the entire environment.
2. Some converged infrastructure solutions are still pretty new.
Yes, there’s a lot of buzz around converged infrastructure, but the technology – in some cases - is still relatively new—at least as it is being implemented today. As the design of converged infrastructure progresses, support for different configuration models will expand. This technology is designed to enable optimal scalability--regardless of the underlying hardware or hypervisor--while integrating the power of software-defined storage. Furthermore, a converged architecture helps pool siloed resources and allows them to be truly utilized.
As with any new platform, however, there may be hiccups. Be aware of new configurations around pooled resources and how they’re allocated. You’re now working with a non-traditional storage mechanism, so it’s critical to validate the workloads that will live on a converged infrastructure environment prior to going live. A good performance assessment will stress-test user loads against specific repositories. This will give you a clear benchmark for performance and user load.
3. Converged architecture redefines storage and infrastructure management.
The best way to really understand converged infrastructure is to see the bigger picture.
You may currently have several storage and compute silos in your organization. Converged architecture systems are designed to eliminate those silos within the data center and within the cloud.
Indeed, converged architecture allows organizations to deploy several new kinds of storage and compute models. For example, if you have an older array of disks, you can point them to the converged architecture controller to give it new life.
Another important point to consider is the amount of control converged architecture systems give organizations around physical media. Traditional disks, hybrid arrays and all-flash arrays can be controlled and optimized through a converged architecture. At this point, you can create intelligent policies that can dynamically allocate workloads to appropriate disks. This can be assigned to applications, databases, backup operations, cloud extensions and more. Remember, the idea isn’t just to create a “converged” architecture; you’re also trying to create an environment that’s easier to manage and is built around intelligence.
New types of converged infrastructure systems are helping define the next-generation data center. Organizations looking to create cloud-scale ecosystems must look to next-generation converged architecture to help them evolve.
This content is underwritten by Hewlett Packard Enterprise (HPE).
