So you think you’re ready for your journey to the cloud. Before you hop in and shift gears, there’s a little engine maintenance in order.
As the amount of data, applications and servers in data centers have increases, too have corresponding storage and networks. One of the challenges of storage-specific networking protocols is that they’re incompatible with the dominant server networking protocol: Ethernet. Organizations have therefore had to deploy entirely separate physical networks in their data centers: an Ethernet LAN for server connectivity and a SAN for storage. Although the two mostly perform the same function, they have separate physical networks, with separate switches, cabling, networking hardware and connections into each server. This creates a lot of potential complexity that is an easy target for elimination.
Your data center has gotten crowded.
Because these networks use different protocols, SANs and Ethernet varieties often need to be maintained by separate teams with different skill sets. Not only is this inefficient because it requires potentially redundant network management staff, now adding, moving, or changing physical servers and connections needs coordination between two groups. That makes it costly and inefficient for organizations and reduces their flexibility to adapt to changing business needs – like getting into the cloud.
One answer to this housekeeping challenge has been a new standard called Data Center Bridging (DCBx), an architecture based on a collection of open standards Ethernet extensions developed and ratified by the IEEE 802.1x working group, which was tasked with improving and expanding networking and management capabilities in the data center. DCBx collapses networks down to a single common network infrastructure and unifies all these disparate networking fabrics – eliminating redundant switches, cables and while providing a “wire once, provision many” type of architecture.
This Unified Fabric Networking is really about the fabrics of connectivity on the data center floor so that as many technologies as possible work through one single wire. Now you’re able to provision any connectivity you want across that wire. No longer do you have to bring down a machine and re-cable or bring in more switches. It can all be done elastically and dynamically.
With Unified Fabric Networking technologies, data centers are seeing:
Efficiency: By eliminating a large amount of infrastructure redundancy across the network rather than just at the access layer, data centers see 80% less cable, 8 times better performance and 20% of the infrastructure of traditional environments – not to mention simplified administration and maintenance.
Agility: Provides the ability to set up, move, and change both physical and virtual servers faster to respond to changing business needs. The approach is designed to provide the ability to adjust based on capacity needs – regardless of protocol.
IT transformation: By providing a simpler, more homogeneous infrastructure to manage and support a capacity demand model, Unified Fabric Networking technologies can help IT organizations do more with less.
In fact, a recent IDC study analyzing the experience of six companies that have undergone implementations of this type of network convergence tune-up estimated that these businesses were able to achieve up to a 492% ROI by fully converging their network at both the access and storage layers.1
It’s the kind of bottom-line benefit every organization wants. But first it means taking a good, honest look at what’s under the hood in your data center.