Distributed Software-Defined Networking: The ACM PODC 2014 Workshop DSDN

نویسندگان

  • Petr Kuznetsov
  • Stefan Schmid
چکیده

The workshop on Distributed Software-Defined Networking, DSDN, took place in Paris, France, on the 15th of July, just before the 33rd ACM Symposium on Principles of Distributed Computing. The workshop intended to be a forum to discuss new algorithmic and distributed computing challenges offered by the emerging field of Software-Defined Networking (SDN). The workshop consisted of invited and peer-reviewed presentations, both from researchers in the field of distributed computing and in the field of networking. 1 SDN: Networking Is Cool Again! Computer networking currently goes through a transition phase, and the paradigm of SoftwareDefined Networking (SDN) is discussed intensively, both in the industry and in the academia. In a nutshell, the paradigm out-sources and consolidates the control over a network to a logically centralized software control plane. This separation, and the introduction of “programmability”, allows to adapt and innovate the network control plane more efficiently, and independently of the data plane. The resulting flexibilities open interesting new opportunities. Or, as the Cisco CTO recently put it: networking is cool again! 2 SDN meets PODC At the heart of SDN lies the idea to design and operate the network control logic on a centralized network view. However, inevitably, this view is only logically centralized [5, 6, 8, 10]: in order to avoid a single point of failure and ensure scalability and efficiency, the control plane state must be physically distributed. The design of a distributed control plane is only one example where we feel that the PODC community could contribute to relevant networking problems today. Accordingly, the goal of the DSDN workshop was to bring together networking researchers with the PODC community, to discuss current trends in networking, and to identify interesting articulation points between the two communities. The workshop took place in Paris on the sunny day of July 15th, just before ACM PODC, concurrently with four other PODC workshops. 3 Program Below we give a short summary of the talks given in our workshop. Abstracts and slides can be found at http://www.podc.org/podc2014/dsdn14/. 3.1 Foundations of SDN The opening keynote was given by Nate Foster (Cornell University), one of the leading figures in SDN and whose research is situated at the intersection of programming languages, networks, and security. Nate gave an overview of the motivation for and foundations of SDN, and emphasized the importance that programming languages and formal methods play in software-defined networks: Only by a careful engineering of the right programming abstractions, an effective reasoning about network behavior becomes possible. The ability to formally reason about an SDN system is crucial, especially in the light of today’s trend towards public cloud computing, where a network misconfiguration may leak confidential information to other tenants. Nate presented his vision of the machine, language, and runtime models for SDN. In particular, he described the design of a machine-verified SDN controller, which is based on a detailed operational OpenFlow model and which is formalized in Coq. [4] This operational OpenFlow model can also be used to develop a verified compiler and run-time system for a more high-level network programming language such as NetKAT. [2] 3.2 Consistent Range Classification with OpenFlow Yehuda Afek (Tel Aviv University) presented his recent work with Anat Bremler-Barr and Liron Schiff [1] on consistently managing flows and classifying ranges in SDN networks with multiple entrance, i.e., in scenarios where the flow changes the entrance point to the network. Their range classification scheme only requires three entries per range, and supports atomic updates across multiple switches. This makes the scheme attractive, e.g., in load-balancing and NFV applications. 3.3 SDN-Based Private Interconnection Shlomi Dolev presented his joint work with Shimrit Tzur-David. Motivated by the advent of hybrid clouds, clouds augmenting private datacenters with the public clouds, Shlomi Dolev presented a new approach for private communication and data transfers leveraged by SDN. The work assumes that SDN can enable deterministic and manageable private virtual networks between the local datacenters that reside in the private cloud, to the public resources in the public cloud. In particular, Shlomi presented a private hybrid cloud in which all the information that passes across the cloud is information-theoretically secure, i.e., unless there is a sufficiently large coalition of malicious routers, the information cannot be revealed. The main idea of the approach is to use a secret sharing scheme together with SDN, to ensure privacy over multipath communication. 3.4 Software Transactional Networking: A Robust and Distributed SDN Control Plane Marco Canini (Université Catholique de Louvain) presented the concept of software transactional networking (STN), a control-plane abstraction used for consistent composition of concurrent poli-

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Work and Publications

s: Some of the original abstracts are reproduced almost without change and some with minor revision. The former cases are indicated by v.o., whereas the latter cases are indicated by rev. Common abbreviations (for the publication items) include: CCC: Annual IEEE Conference on Computational Complexity. COLT: Annual ACM Workshop on Computational Learning Theory. FOCS: Annual IEEE Symposium on Fou...

متن کامل

The 2010 Edsger W. Dijkstra Prize in Distributed Computing

The ACM-EATCS Edsger W. Dijkstra Prize in Distributed Computing was created to acknowledge outstanding papers on the principles of distributed computing whose significance and impact on the theory or practice of distributed computing have been evident for at least a decade. The Prize is sponsored jointly by the ACM Symposium on Principles of Distributed Computing (PODC) and the EATCS Symposium ...

متن کامل

Vers l'auto-stabilisation des systèmes à grande échelle. (Toward self-stabilizing large-scale systems)

. In Gerard Tel and PaulM. B. Vitányi, editors,Distributed Al-gorithms, 8th International Workshop, WDAG ’94, volume 857 of LectureNotes in Computer Science, pages 326–339. Springer, 1994. [11] Joffroy Beauquier, Béatrice Bérard, Laurent Fribourg, and FrédéricMagniette. Proving convergence of self-stabilizing systems usingfirst-order rewriting and regular languages. Distributed ...

متن کامل

Infinite Networks, Halting and Local Algorithms

The immediate past has witnessed an increased amount of interest in local algorithms, i.e., constant time distributed algorithms. In a recent survey of the topic (Suomela, ACM Computing Surveys, 2013), it is argued that local algorithms provide a natural framework that could be used in order to theoretically control infinite networks in finite time. We study a comprehensive collection of distri...

متن کامل

Message from the ManSDN/NFV workshop co-chairs

Welcome to the second International Workshop on Management of SDN and NFV Systems (ManSDN/NFV 2015)! Following a successful event in 2014, we are pleased to organize the workshop this year also and provide a venue for discussions on SDN and NFV management issues. We hope that, with its high-quality technical program and its pleasant setting, the workshop will stimulate profitable interaction am...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2014