CRII: NeTS: A Coordinate System Based Network for Data Plane Scalability and High Throughput

Increasingly many application of cloud service and big data processing requires the design of a high- throughput network that provides scalability for tens of thousands of servers and flexibility for incremental growth of network size. Recent work has been investigating new network architecture and protocols to provide the three important requirements of data centers, namely, high throughput, flexibility for incremental growth, and data plane scalability. Unfortunately, existing data center network architectures focus on one or two of the above properties and pay little attention to the others. This research proposes to systematically study the architecture and design issues of a novel coordinate- based network architecture that meets all three requirements of data center network design. Our design innovations include (i) a random and dense interconnect of switches to achieve flexibility, (ii) a novel greedy routing protocol on virtual coordinates to achieve high throughput, and (iii) using virtual coordinates for routing instead of host addresses to achieve data plane scalability. In addition, we also propose to allow the network to be powerful to support various data center services such as key-based operations, multicast, and traffic monitoring. Compared to existing work, the coordinate- based network provide new opportunities to achieve these services without sacrificing data plane scalability. The research objectives include the following: (i) design, implementation, and evaluation of a coordinate- based network architecture; (ii) improving the network throughput by introducing load-aware routing mech- anism; (iii) design and implementation of scalable data center services in the proposed network, including key-value operations, multicast, and traffic monitoring; and (iv) development of a unified simulator of data center network protocols. Intellectual Merits: The intellectual merits of this research consist of three areas: (i) the development of new protocols and techniques that improve data center network performance; (ii) providing design con- siderations to computer networking community to seek new architectures which meet various application requirements; (iii) an algorithmic foundation of the design and analysis of greedy routing on various network topologies. Broader Impacts: The results of this research contribute towards enhancing the performance and s- calability of current data center networks, the creation of a novel greedy routing protocol using virtual coordinates, and the education of undergraduate and graduate students on network protocol design. Giv- en the growing importance of big data applications and cloud computing, the proposed research has the potential to influence network operation in academia, industry, business, government, etc. The results of this research will be incorporated into the undergraduate course Computer Networking and Introduction to Database Systems and graduate courses Advanced Computer Networking taught by the PI at the University of Kentucky. Students trained in this project will be capable to develop new protocols and conduct network experiments in various test tools. This project will help the state of Kentucky to strengthen the capacity of researchers and improve the quality of research and development. The PI is also making effort to integrate under-represented minorities. Finally, the simulator built for this work will be disseminated in the form of an open-source prototype.