Fabric Networking
We all know the wise maxim, "Less is more," and it is finally true of networks. Ethernet fabrics decrease the number of devices on the network, as well as the number of tiers to create flatter, faster networks that preserve the low cost and simplicity of Ethernet. This innovative network advancement takes the most prized qualities of Ethernet, and enhances them for today's virtualised data center and changing business requirements. Ethernet fabric-based networks address data center pain points by providing:
- Automation
- Reduced network complexity
- Tighter Virtual Machine (VM) integration and seamless VM mobility
- Simplified management
- Increased scalability
- Improved performance
Ethernet Fabrics In Ethernet fabrics, the large, flat Layer 2 network delivers high wire-speed performance and high network resiliency. In addition, all paths between switches are fully active because the network does not rely on STP. The network topology is flexible and dynamic—changing quickly as the needs of the business change. And, if appropriate to the application, IP and storage traffic can be converged over a common network infrastructure, further reducing cost. Ethernet fabrics enable intelligent and seamless VM mobility and simplified management across the entire data center environment. All devices in the fabric are aware of each other. As a VM moves, manual reconfiguration is no longer needed because the VM's profile information already exists in the fabric and is known by all network devices. Network administration is simplified since all switches in the fabric can be managed as a single entity or individually as needed. Last, Ethernet fabrics are self-forming, self-aggregating, and VM-aware. By simply having an administrator add a switch to the fabric, Inter-Switch Links (ISLs) are automatically configured and aggregated, and VM profile settings automatically extracted from the hypervisor and applied. Adding value– the move to the cloud Once an Ethernet fabric-based infrastructure is established, the next step is to begin moving toward a private cloud model. In this stage, data centers will have a large, flat, fully utilised Layer 2 network that provides high bandwidth and delivers a high level of network automation. In some cases— and for some applications—IT can begin connecting storage resources to the Ethernet fabric. With this infrastructure in place, IT organisations are able to scale virtualisation, improve and refine automation, and rewrite IT policies and processes for a more services-oriented approach. Extended Private Clouds To create an extended private cloud and seamlessly extend VM mobility to other data centers, organisations will need to extend the Layer 2 network over distance and be able to accelerate and access applications over WAN connections. Using Ethernet fabrics as the foundation, the extended private cloud will allow IT organisations to expand their pools of compute, storage, and network resources by leveraging infrastructure investments across multiple data center locations. IT will gain more flexibility to scale applications and meet rapid increases in demand. Hybrid Clouds The hybrid cloud builds on to the concepts of the private and extended private cloud. A hybrid cloud model, with its Ethernet fabrics, requires a network infrastructure built with standards-based technology; open support for hypervisors, servers, and storage; and tight integration with systems management applications. In a hybrid cloud, resources from the private and public clouds are combined so that businesses can be more agile and responsive. The hybrid cloud allows IT organisations to leverage other data center locations and resources at service providers in the public cloud. Applications, such as e-mail, data storage, and Customer Relationship Management (CRM), will be served through the cloud at lower costs. This model gives IT organisations more options to cost-effectively meet spikes or seasonal demand for applications and storage, supports speedy application deployments, and ensures high service levels in the event of a local outage.
