What matters most in creating true global IT efficiency?
When I speak with people about “global efficiency,” they generally assume we’re going to start off by talking about the storage system (we do sell storage, after all), and zero in on product features like deduplication, compression, and inline efficiency. Those are great things to talk about, and a very interesting part of the conversation, but they are far from the full picture of “global efficiency” in IT infrastructure.
To get a more open forum conversation going around this key element in the next generation data center, I recently created a whiteboard video talking through some of the lesser-discussed, but equally as (if not more) important points.
To achieve global efficiency, consolidation is required to make everything simpler and more common through all areas of the infrastructure.
The obvious parts of consolidation include:
- Standardizing the network layer and how it is controlled
- Consolidating storage platforms into one that is scalable and flexible
- Combining design and operations teams
- Making compute a single fungible resource
The infrastructure spaghetti monster
Driving a mindset that has set IT organizations on a quest for better global efficiency are virtualization environments that have sprawled to a point at which the complexity of it all can’t be fixed or “hidden” by automation tasks. By only looking at individual pieces of architecture across several operations teams, the whole system ends up as one big spaghetti infrastructure. This leaves operations dealing with multiple server platforms, disparate networks, storage protocols, and vendors. In this mode, the complexity itself becomes a burden.
Due to this pain, forward-thinking CIOs are looking at their next generation data center as a global project that must simplify everything. One team working together on all infrastructure, fewer systems, fewer interfaces, and more commonality between components are all key pillars to building for global efficiency.
What matters most in creating true global efficiency?
Low initial capital is not the highest priority when you’re desperate to move away from the spaghetti mess and into an architecture you can claim is truly globally efficient — and that can feel counter-intuitive to those who are not yet working at scale. But building whatever was cheapest at the time was how so many got into so much pain in the first place.
What should be taken into account more primarily is the number of use cases that can be consolidated onto a single platform. The ideal scenario is having one infrastructure design that encompasses all compute, networking, and storage — for all use cases. A true Infrastructure as a Service (IaaS).
Of course, it does matter how much that infrastructure costs to build and maintain. But for those looking to achieve a true next generation data center, you must put simplification and consolidation first.
With this lense on design, it’s understandable why heads are turning toward significantly different platforms than what the industry has offered historically. Systems that can be flexible, large-scale, high density, and run in standardized protocols are now preferred. Part of the desire for an SDDC is having the ability to build a pool of resources connected across standard ethernet, and define any infrastructure required in that standard pool. This has heavily influenced our product designs at SolidFire. It’s why a shared-nothing, ethernet-based, commodity hardware is our preferred path.
We are helping customers move to a truly globally efficient next generation data center by providing the foundation to build that simplified, consolidated design.
Watch the whiteboard video on global efficiency in the data center, or read more of our posts on the Next Generation Data Center.
| 0 comments
Posted in Next Generation Data Center.