NaaS: Network-as-a-Service in the Cloud

Cloud computing has significantly changed the IT landscape. Today it is possible for small companies or even single individuals to access virtually unlimited resources in large data centres (DCs) for running computationally demanding tasks. This has triggered the rise of "big data" applications, which operate on large amounts of data. These include traditional batch-oriented applications, such as data mining, data indexing, log collection and analysis, and scientific applications, as well as real-time stream processing, web search and advertising.

To support big data applications, parallel processing systems, such as MapReduce, adopt a partition/aggregate model: a large input data set is distributed over many servers, and each server processes a share of the data. Locally generated intermediate results must then be aggregated to obtain the final result.

An open challenge of the partition/aggregate model is that it results in high contention for network resources in DCs when a large amount of data traffic is exchanged between servers. Facebook reports that, for 26% of processing tasks, network transfers are responsible for more than 50% of the execution time. This is consistent with other studies, showing that the network is often the bottleneck in big data applications.

Improving the performance of such network-bound applications in DCs has attracted much interest from the research community. A class of solutions focuses on reducing bandwidth usage by employing overlay networks to distribute data and to perform partial aggregation. However, this requires applications to reverse-engineer the physical network topology to optimise the layout of overlay networks. Even with perfect knowledge of the physical topology, there are still fundamental inefficiencies: e.g. any logical topology with a server fan-out higher than one cannot be mapped optimally to the physical network if servers have only a single network interface.

Other proposals increase network bandwidth through more complex topologies or higher-capacity networks. New topologies and network over-provisioning, however, increase the DC operational and capital expenditures-up to 5 times according to some estimates-which directly impacts tenant costs. For example, Amazon AWS recently introduced Cluster Compute instances with full-bisection 10 Gbps bandwidth, with an hourly cost of 16 times the default.

In contrast, we argue that the problem can be solved more effectively by providing DC tenants with efficient, easy and safe control of network operations. Instead of over-provisioning, we focus on optimising network traffic by exploiting application-specific knowledge. We term this approach "network-as-a-service" (NaaS) because it allows tenants to customise the service that they receive from the network.
NaaS-enabled tenants can deploy custom routing protocols, including multicast services or anycast/incast protocols, as well as more sophisticated mechanisms, such as content-based routing and content-centric networking.

By modifying the content of packets on-path, they can efficiently implement advanced, application-specific network services, such as in-network data aggregation and smart caching. Parallel processing systems such as MapReduce would greatly benefit because data can be aggregated on-path, thus reducing execution times. Key-value stores (e.g. memcached) can improve their performance by caching popular keys within the network, which decreases latency and bandwidth usage compared to end-host-only deployments.

The NaaS model has the potential to revolutionise current cloud computing offerings by increasing the performance of tenants' applications -through efficient in-network processing- while reducing development complexity. It aims to combine distributed computation and network communication in a single, coherent abstraction, providing a significant step towards the vision of "the DC is the computer".

The importance of cloud computing for future services and applications has been recognised widely. A recent report by IBISWorld predicts that the UK cloud computing market will grow at an annual rate of 15.8% from £ 5.2bn in 2011/2012 to reach £ 11bn by 2016/2017. As well as decreasing capital and operational expenditure through outsourcing ICT infrastructure, cloud computing potentially reduces time-to-market. This creates numerous opportunities for SMEs and public organisations, thus engaging with providers of public, private or community clouds.

The NaaS project formulates an innovative new infrastructure facility for cloud data centres with the primary goal of improving the performance and cost effectiveness of network-bound cloud applications for the benefit of end
users. As such, the path to impact for the project outputs runs directly through data centre operators and application developers. For this reason, our impact plan focuses on reaching and influencing two main constituencies:
(1) industrial companies providing data centre infrastructure platforms, services and applications and (2) standards bodies developing and promoting open standards in the cloud data centre arena.

The challenges of cloud computing cut a remarkably wide swath through industry. The letters of support accompanying this proposal indicate the serious and deep interest in the NaaS project from a broad array of companies:
Xilinx is the world leader in programmable hardware accelerators; Netronome is a rapidly growing UK-based provider of high-speed network packet processors; Citrix, through its UK-developed Xen suite of products, is the world leader in virtualisation technology for cloud data centres; and NetApp is a leading provider of storage and data management solutions for cloud-hosted, data-intensive applications. These companies, which represent the key elements of any cloud data centre technology stack, are eager to provide support and gain the outputs of our research for insertion into the competitive cloud marketplace.

We expect the NaaS project to contribute significantly to open standards relating to programmable networking and cloud computing platforms such as the Open Networking Foundation and the Open Source Routing Forum. Standardisation efforts in software-defined networking (SDN) and cloud computing are highly prominent at the moment and experience broad commercial and academic support.

The open source NetFPGA platform is of particular relevance and utility to the NaaS project. The NetFPGA platform is a joint project between Stanford University and Cambridge University, providing an ideal means to disseminate NaaS technology developments to a wide community of researchers, adopters and investors. We have previously used it as a vehicle for disseminating our ideas in OpenFlow and SDN technology.

To stimulate a quick uptake of NaaS technologies by the research community, as well as interested industrial and government agencies, we will make prototypes available under an open source licence. Our goal is to build an active user group in order to eventually provide the research community and industry with mature instruments for exploring theoretical and practical aspects of the NaaS vision.

The novelty, timeliness and relevance of the NaaS project will ensure widespread interest in its results from the international community. There will also be abundant opportunities for the technology to be exploited by the
industrial supporters. Through those supporters, the NaaS project will provide significant benefit to the UK software industry, increasing its competitiveness in the growing global marketplace for cloud data centres and services. For the academic partners, the project will lead to high-profile publications and to exciting opportunities for future research. This will further the UK's aspirations for the development and use of cloud computing.