Workshop titled "Toward a Research Agenda for Cloud 3.0"

University Of Wisconsin-Madison, Madison WI

Investigators

Abstract

This award funds a focused workshop with the goal of bringing together leading researchers and practitioners in networked systems to develop a new networking research agenda for "Cloud 3.0", focusing on rethinking from the ground-up networked systems infrastructure within the datacenter and beyond. Background: In recent years we have witnessed two dramatic transformations of the IT industry. First, virtualization, and private data centers, enabled organizations to use limited compute resources in a more flexible and cost-effective fashion. Second, cloud computing, and public clouds, enabled organizations to outsource their IT infrastructure, vastly reducing capital and operational expenses. We are now poised for the third dramatic shift: "serverless" platforms, or "Cloud 3.0", which has the potential to reinvent both hardware and software architectures. With Cloud 3.0, application developers just bring code and data without worrying at all about the infrastructure (e.g., configuring and spinning up cloud resources). Application developers define even the most complex applications as a set of simple stateless functions, or lambdas. An infrastructure provider (e.g., a public cloud operator) handles all the complex "details" (e.g., acquiring resources, managing failures, ensure good user experience, etc.). This model offers many benefits including development of innovative applications at much greater speed, rapid adaptation to user demand, improved cost-effectiveness, etc. Several large online service providers have begun offering platforms for serverless computation, touting this as "the future of the Cloud". The current approach to supporting such Cloud 3.0 applications, infrastructure, and networking is to re-purpose existing hardware and software, protocols, and algorithms. But there are several places, especially in cloud networking, where existing technologies are a poor fit for Cloud 3.0. For example, at the hardware level, recent research in disaggregated network designs has the potential to further erode the hard boundaries defining servers as well. Similarly, new approaches to rethinking network protocols, abstractions, application programming interfaces (APIs), and algorithms may lead to more optimal support for serverless computing. "Cloud 3.0" provides a rich set of opportunities for interesting core networking, as well as cross-cutting ("networking + X") research. Topics of interest that are likely to be explored at the workshop include: new datacenter architectures and networking interconnects; synergy between Lambdas and disaggregated network architectures; routing and link layer protocols to support Lambdas? performance, scale up, isolation; programmable fabrics and control planes that can support agility needed by Lambdas; revisiting transport and network layer protocol design; revisiting old abstractions such as the socket A ; Lambda isolation, security, and accounting, and the necessary support at network. The principal investigator, along with workshop participants, will author a detailed workshop report which will outline a networking research agenda for Cloud 3.0. The report will be made available on the public workshop website. The website will also include a list of participants, their biographies, and their one-page pre-workshop write-ups.