Today’s companies are becoming more and more reliant on big data to deliver better products and services to market. As part of improving competitiveness and serving customers better, the reliance on bigger and faster data centres grows each day. The modern data hungry businesses of tomorrow can no longer rely on legacy Ethernet data centre infrastructures in use today.
While cloud computing and virtualisation has helped to address increasing data demands, network topologies have remained the same. Now, Ethernet fabrics have been designed to overcome the issues of legacy Ethernet networks and address the change in traffic patterns created by virtualised data centre environments.
Data centres that work
Traditional data centres were designed to accommodate the traffic patterns of client-server computing environments. This ‘North-South’ flow saw data predominantly traversing between clients and servers with little traffic directly between the servers. Modern data centre traffic on the other hand is affected by the heavy use of virtualisation within the server environment, changing the flow in an ‘East-West’ direction with much of the data being between servers rather than between clients and servers. Ethernet fabric pre-determines the route that traffic will take ensuring the shortest path to the destination. This improves latency and simplifies traffic routing helping to decrease costs.
In today’s legacy data centre networks providing resilience typically results in the deployment of large numbers of redundant and costly links. Legacy protocols do not allow for efficient use of these redundant links - with links either being used in an active-standby mode or poorly utilised with uneven load sharing in an active-active mode. Ethernet fabrics are highly efficient and obtaining almost 100% link utilisation is possible. This high level of resilience comes with support for link aggregation group termination across multiple switch devices within the fabric.
Ethernet Fabrics offer businesses a number of benefits:
- A flatter topology reduces complexity whilst providing deterministic latency
- Flexible architectures provide multiple potential topologies including; full mesh, partial mesh, Clos and hybrid providing you with the choice and flexibility to employ the architecture that works for your environment
- Plug and play capability provides a simple scalable and resilient architecture without the need for legacy protocols such as Spanning Tree Protocol (STP), reducing costs and simplifying potentially complex envirornments
- Multiple, resilient paths provide high performance and resilience
Redcentric’s Ethernet fabric
Redcentric specialises in the design, build and deployment of Ethernet fabric networks to support today’s data centre requirements and traffic patterns. We adopt a consultative approach to our design process with a strong emphasis on interaction with your technical staff to ensure we understand the unique requirements of your specific environment.
Why Redcentric for Ethernet Fabric?
- Vendor neutral – we supply Ethernet fabric networks from the world’s leading providers, including both Brocade and Cisco.
- Experience – we have over five years of experience in building Ethernet fabric based data centre networks in a variety of complex environments.
- Capability – we have already designed and built many Ethernet fabric based data centre networks for mission critical infrastructures that support 24/7/365 operations and provide up to 99.999% availability.
- Support – we have extensive 24/7 UK wide support capability, and offer a range of SLAs available to suit mission critical and not so critical networks.