CCR Papers from July 2015

  • Dina Papagiannaki

    Welcome to the July issue of Computer Communications Review. Over the past months we have tried to make CCR a resource where our community publishes fresh new ideas, expresses positions on interesting new research directions, as well as reports back on community activities, like workshops and regional meetings. In addition, we have introduced the student advice column and the column edited by our industrial board aiming to provide a clearer bridge between scientific practice and technology in commercial products. I am really proud to see all these additions being embraced by the community.

    This issue features three technical papers, two on cloud computing and one on TCP. Our sincerest thanks to all the authors that submit technical contributions to CCR. And my personal thanks to the tireless editorial board that always aims towards outstanding quality while providing constructive feedback for the continuous improvement of the submitted manuscripts.

    The editorial zone features three papers as well. One reports on the workshop on Internet economics that took place late last year. The other two cover (i) research challenges in multi-domain network measurement and monitoring, and (ii) the lessons learnt from using RIPE’s ATLAS for measurement research. I hope that both editorials will provide useful insights to the CCR audience.

    Our industrial column features a contribution from Akamai. The authors describe how research in our community has influenced the design of the content delivery network (CDN) of Akamai, as well as the practical realities they had to deal with to create that bridge between academic output and a scalable, robust content delivery network that serves trillion of requests per day.

    This issue is coming one month before the annual SIGCOMM conference, that will take place in London in August. Since a couple of years ago, SIGCOMM is the venue where CCR is presenting its "best-of" papers for the previous four issues (July 2014 to April 2015). I am really happy to announce the following two "best of CCR" papers for the year 2014-2015.

    Technical paper: "Programming Protocol-Independent Packet Processors", by P. Bosshart (Barefoot Networks), D. Daly (Intel), G. Gibb (Barefoot Networks), M. Izzard (Barefoot Networks), N. McKeown (Stanford University), J. Rexford (Princeton University), C. Schlesinger (Barefoot Networks), D. Talayco (Barefoot Networks), A. Vahdat (Google), G. Varghese (Microsoft), and D. Walker (Princeton University).

    Editorial paper: "A Primer on IPv4 Scarcity", by P. Richter (TU Berlin/ICSI), M. Allman (ICSI), R. Bush (Internet Initiative Japan), and V. Paxson (UC Berkeley/ICSI).

    Congratulations to the authors of the two best papers and I am really looking forward to seeing all of you in London in August. Dina Papagiannaki CCR Editor

  • S. Yaw, E. Howard, B. Mumey, M. Wittie

    Given a set of datacenters and groups of application clients, well-connected datacenters can be rented as traffic proxies to reduce client latency. Rental costs must be minimized while meeting the application specific latency needs. Here, we formally define the Cooperative Group Provisioning problem and show it is NP-hard to approximate within a constant factor. We introduce a novel greedy approach and demonstrate its promise through extensive simulation using real cloud network topology measurements and realistic client churn. We find that multi-cloud deployments dramatically increase the likelihood of meeting group latency thresholds with minimal cost increase compared to single-cloud deployments.

    Phillipa Gill
  • C. Fuerst, M. Rost, S. Schmid

    It is well-known that cloud application performance can critically depend on the network. Over the last years, several systems have been developed which provide the application with the illusion of a virtual cluster : a star-shaped virtual network topology connecting virtual machines to a logical switch with absolute bandwidth guarantees. In this paper, we debunk some of the myths around the virtual cluster embedding problem. First, we show that the virtual cluster embedding problem is not NP-hard, and present the fast and optimal embedding algorithm VC-ACE for arbitrary datacenter topologies. Second, we argue that resources may be wasted by enforcing star-topology embeddings, and alternatively promote a hose embedding approach. We discuss the computational complexity of hose embeddings and derive the HVC-ACE algorithm. Using simulations we substantiate the benefits of hose embeddings in terms of acceptance ratio and resource footprint.

    Hitesh Ballani
  • H. Ding, M. Rabinovich

    This paper examines several TCP characteristics and their effect on existing passive RTT measurement techniques. In particular, using packet traces from three geographically distributed vantage points, we find relatively low use of TCP timestamps and significant presence of stretch acknowledgements. While the former simply affects the applicability of some measurement techniques, the latter may in principle affect the accuracy of RTT estimation. Using these insights, we quantify implications of common methodologies for passive RTT measurement. In particular, we show that, unlike delayed TCP acknowledgement, stretch acknowledgements do not distort RTT estimations.

    Joseph Camp
  • P. Calyam, M. Swany

    The perfSONAR-based Multi-domain Network Performance Measurement and Monitoring Workshop was held on February 20-21, 2014 in Arlington, VA. The goal of the workshop was to review the state of the perfSONAR effort and catalyze future directions by cross-fertilizing ideas, and distilling common themes among the diverse perfSONAR stakeholders that include: network operators and managers, endusers and network researchers. The timing and organization for the second workshop is significant because there are an increasing number of groups within NSF supported data-intensive computing and networking programs that are dealing with measurement, monitoring and troubleshooting of multi-domain issues. These groups are forming explicit measurement federations using perfSONAR to address a wide range of issues. In addition, the emergence and wide-adoption of new paradigms such as software-defined networking are taking shape to aid in traffic management needs of scientific communities and network operators. Consequently, there are new challenges that need to be addressed for extensible and programmable instrumentation, measurement data analysis, visualization and middleware security features in perfSONAR. This report summarizes the workshop efforts to bring together diverse groups for delivering targeted short/long talks, sharing latest advances, and identifying gaps that exist in the community for solving end-toend performance problems in an effective, scalable fashion.

  • V. Bajpai, S. Eravuchira, J. Schönwälder

    We reflect upon our experience in using the RIPE Atlas platform for measurement-based research. We show how in addition to credits, control checks using rate limits are in place to ensure that the platform does not get overloaded with measurements. We show how the Autonomous System (AS)-based distribution of RIPE Atlas probes is heavily skewed which limits possibilities of measurements sourced from a specific origin-AS. We discuss the significance of probe calibration and how we leverage it to identify load issues in older hardware versions (38.6% overall as of Sep 2014) of probes. We show how performance measurement platforms (such as RIPE Atlas, SamKnows, BISmark and Dasu) can benefit from each other by demonstrating two example use-cases. We also open discussion on how RIPE Atlas deployment can be made more useful by relaying more probe metadata information back to the scientific community and by strategically deploying probes to reduce the inherent sampling bias embedded in probe-based measurement platforms.

  • kc claffy, D. Clark

    On December 10-11 2014, we hosted the 4th interdisciplinary Workshop on Internet Economics (WIE) at the UC San Diego’s Supercomputer Center. This workshop series provides a forum for researchers, Internet facilities and service providers, technologists, economists, theorists, policy makers, and other stakeholders to inform current and emerging regulatory and policy debates. The objective for this year’s workshop was a structured consideration of whether and how policy-makers should try to shape the future of the Internet. To structure the discussion about policy, we began the workshop with a list of potential aspirations for our future telecommunications infrastructure (a list we had previously collated), and asked participants to articulate an aspiration or fear they had about the future of the Internet, which we summarized and discussed on the second day. The focus on aspirations was motivated by the high-level observation that before discussing regulation, we must agree on the objective of the regulation, and why the intended outcome is justified. In parallel, we used a similar format as in previous years: a series of focused sessions, where 3-4 presenters each prepared 10-minute talks on issues in recent regulatory discourse, followed by in-depth discussions. This report highlights the discussions and presents relevant open research questions identified by participants. Slides presented and a copy of this report are available at http://www.caida.org/workshops/wie/1412/.

  • A. Akella

    I've received several interesting, and varied, questions from students all over the world. Thank you for the warm response! In this issue, I have hand-picked a small subset of questions to answer. Many thanks to Brighten Godfrey (UIUC) and Vyas Sekar (CMU) for contributing their thoughts.

  • B. Davie, C. Diot, L. Eggert, N. McKeown, V. Padmanabhan, R. Teixeira

    As networking researchers, we love to work on ideas that improve the practice of networking. In the early pioneering days of the Internet the link between networking researchers and practitioners was strong; the community was small and everyone knew each other. Not only were there many important ideas from the research community that affected the practice of networking, we were all very aware of them. Today, the networking industry is enormous and the practice of networking spans many network equipment vendors, operators, chip builders, the IETF, data centers, wireless and cellular, and so on. There continue to be many transfers of ideas, but there isn’t a forum to learn about them. The goal of this series is to create such a forum by presenting articles that shine a spotlight on specific examples; not only on the technology and ideas, but also on the path the ideas took to affect the practice. Sometimes a research paper was picked up by chance; but more often, the researchers worked hand-in-hand with the standards community, the open-source community or industry to develop the idea further to make it suitable for adoption.

  • B. Maggs, R. Sitaraman

    This paper peeks under the covers at the subsystems that provide the basic functionality of a leading content delivery network. Based on our experiences in building one of the largest distributed systems in the world, we illustrate how sophisticated algorithmic research has been adapted to balance the load between and within server clusters, manage the caches on servers, select paths through an overlay routing network, and elect leaders in various contexts. In each instance, we first explain the theory underlying the algorithms, then introduce practical considerations not captured by the theoretical models, and finally describe what is implemented in practice. Through these examples, we highlight the role of algorithmic research in the design of complex networked systems. The paper also illustrates the close synergy that exists between research and industry where research ideas cross over into products and product requirements drive future research.

Syndicate content