Today’s business and government organizations are focused on digital transformation. As a result, dev-ops and business agility is at the forefront of most current strategic IT discussions. To deliver both, enterprise organizations have increasingly leveraged public cloud service providers and are now building on-premise clouds with infrastructure that delivers on three basic requirements:
1. Friction-free agility of physical resources
2. Control systems that maximize physical resource utilization and provide maximum ROI
3. Integration of the various infrastructure components for automated provisioning and resource management
While these requirements exist across the spectrum of infrastructure components (compute, storage, and network), the network plays a foundational role. The network acts as the glue between compute and storage, so the agility, control, and integrality of the network (or lack thereof) directly impacts an organization’s ability to maximize their compute and storage resources. Any bottlenecks in the data center fabric will negatively impact the organization’s ability to run applications efficiently and optimally distribute data. The more the network is in your way, the less business/mission agility can be attained.
Plexxi has created a data center fabric operating environment that leverages commodity off the shelf Ethernet switches and innovative software that forms a single distributed network without the limitations of traditional approaches. For enterprise customers, as well as cloud providers, this modern cloud fabric can be leveraged as a fully integrated data center networking solution (switches, controllers, etc.) for ease of deployment. For larger cloud providers, the constituent technologies are designed to be leveraged as a set of tools woven into the overall cloud orchestration strategy.
Traditional Networks Are Challenged by Today’s Scale-Out Applications
As scale-out converged infrastructure and storage deployments are growing, the data center network is experiencing unprecedented stress and strain from the newly introduced storage traffic and scaled- out applications. Traditional leaf and spine data center architectures, though perform well under normal circumstances, are ill equipped to handle the large volume of highly bursty and unpredictable east/west and north/ south traffic patterns that scale-out applications and storage introduces.
These legacy networ