ABSTRACT
Control of large distributed cloud-based services is a challenging problem. The Distributed Rate Limiting (DRL) paradigm was recently proposed as a mechanism for tackling this problem. The heuristic nature of existing DRL solutions makes their behavior unpredictable and analytically untractable. In this paper we treat the DRL problem in a mathematical framework and propose two novel DRL algorithms that exhibit good and predictable performance. The first algorithm Cloud Control with Constant Probabilities (C3P) solves the DRL problem in best effort environments, emulating the behavior of a single best-effort queue in a fully distributed manner. The second problem we approach is the DRL in processor sharing environments. Our algorithm, Distributed Deficit Round Robin (D2R2), parameterized by parameter α, converges to a state that is, at most, O(1/α) away from the exact emulation of centralized processor sharing queue. The convergence and stability properties are fully analyzed for both C3P and D2R2. Analytical results are validated empirically through a number of representative packet level simulations. The closed-form nature of our results allows simple design rules which, together with extremely low communication overhead, makes the presented algorithms practical and easy to deploy.
- Amazon Simple Storage Service(S3): http://aws.amazon.com/s3.Google Scholar
- A. Berman, R. Plemmons. "Nonnegative matrices in the mathematical sciences". SIAM, 1979.Google Scholar
- D. Bertsekas, R. Gallager. "Data Networks". 1987. Google ScholarDigital Library
- S. Bhatnagar, B. Nath. "Distributed admission control to support guaranteed services in core-stateless networks". In Proceedings of IEEE INFOCOM, 2003.Google Scholar
- S. Boyd, A. Ghosh, B. Prabhakar, D. Shah. "Gossip algorithms: Design, analysis and applications". In Proceedings of IEEE INFOCOM, 2005Google Scholar
- D. F. Carr. "How Google works". Baseline Magazine, July 2006.Google Scholar
- G. Carraro, F. Chong. "Software as a service (SaaS): An enterprise perspective". MSDN Solution Architecture Center, Oct. 2006.Google Scholar
- J. Dilley et al., "Globally Distributed Content Delivery". IEEE Internet Computing, vol. 6(5), 2002. Google ScholarDigital Library
- N. Duffield, P. Goyal, A. Greenberg, P. Mishra, K. Ramakrishnan, J. van der Merive. "A flexible model for resource management in virtual private networks". In Proceedings of ACM SIGCOMM 1999. Google ScholarDigital Library
- V. Dumas, F. Guillemin, P. Robert. "A Markovian analysis of additive-increase multiplicative-decrease algorithms". Adv. in Appl. Probab. 34 (2002), no. 1, 85--111.Google ScholarCross Ref
- R. Gibbens, F. Kelly. "Distributed Connection Acceptance Control for a Connectionless Network", 16th International Teletraffic Conference, Edinburgh, June 1999.Google Scholar
- D. Hinchcliffe. "2007: The year enterprises open thier SOAs to the Internet". Enterprise Web 2.0, Jan. 2007.Google Scholar
- M. Huang. "Planetlab bandwidth limits". Available online: http://www.planet-lab.org/doc/BandwidthLimits.Google Scholar
- A. Jadbabaie, J. Lin, and A. S. Morse, "Coordination of groups of mobile autonomous agents using nearest neighbor rules". IEEE Transactions on Automatic Control, vol. 48(6), 2003Google ScholarCross Ref
- A. Jain, J. M. Hellerstein, S. Ratnasamy, D. Wetherall. "A wakeup call for internet monitoring systems: The case for distributed triggers". In Proceedings of HotNets-III, 2004.Google Scholar
- R. Jain. "The Art of Computer Systems Performance Analysis: Techniques for Experimental Design, Measurement, Simulation, and Modeling". John Wiley and Sons, INC., 1991. Google ScholarDigital Library
- D. Kempe, A. Dobra, J. Gehrke. "Gossip-based computation of aggregate information". In Proceedings of IEEE FOCS, 2003. Google ScholarDigital Library
- C. King, R. Shorten, F. Wirth, M. Akar. "Growth Conditions for the Global stability of Highspeed Communication Networks". To appear in IEEE Transactions on Automatic Control, 2008.Google ScholarCross Ref
- M. Kodialam, T. Lakshman, S. Sengupta. "Maximum Throughput Routing of Traffic in the Hose Model". In Proceedings of IEEE INFOCOM 2006.Google Scholar
- A. Kumar, R. Rastogi, A. Siberschatz, B. Yener. "Algorithms for provisioning virtual private networks in the hose model". IEEE/ACM Trans. on Networking, vol 10(4), 2002. Google ScholarDigital Library
- S. Kunniyur, R. Srikant. "Analysis and Design of an Adaptive Virtual Queue (AVQ) Algorithm for Active Queue Management". IEEE/ACM Trans. on Networking, vol. 12(2). Google ScholarDigital Library
- L. Moreau. "Stability of multiagent systems with time-dependent communication links". IEEE Transactions on Automatic Control, 2005.Google ScholarCross Ref
- A. Odlyzko. "Internet pricing and the history of communications". Computer Networks, vol. 36, 2001.Google Scholar
- J. Padhye, V. Firoiu, D. F. Towsley, J. F. Kurose. "Modeling TCP Reno performance: a simple model and its empirical validation". IEEE/ACM Trans. on Networking, vol 8(2), 2000. Google ScholarDigital Library
- A. Parekh, R. Gallager. "A generalized processor sharing approach to flow control in integrated services networks: the single-node case". IEEE/ACM Trans. on Networking, vol. 1(3). Google ScholarDigital Library
- B. Raghavan, K. Vishwanath, S. Rambhadran, K. Yocum, A. Snoeren. "Cloud Control with Distributed Rate Limiting". In Proceedings of ACM SIGCOMM 2007. Google ScholarDigital Library
- R. Olfati-Saber. "Flocking for multi-agent dynamic systems: algorithms and theory". IEEE Trans. on Auto. Control, 2006.Google ScholarCross Ref
- M. Shreedhar, G. Varghese. "Efficient fair queueing using deficit round-robin". IEEE/ACM Trans. on Networking, 1996. Google ScholarDigital Library
- R. Srikant. "Internet congestion control". Control theory, 14, Birkhauser Boston Inc., Boston, MA, 2004.Google Scholar
- D. Wei, C. Jin, S. Low, S. Hegde. "FAST TCP: motivation, architecture, algorithms, performance". IEEE/ACM Trans. on Networking, 2007. Google ScholarDigital Library
Index Terms
- Fully decentralized emulation of best-effort and processor sharing queues
Recommendations
Cloud control with distributed rate limiting
Today's cloud-based services integrate globally distributed resources into seamless computing platforms. Provisioning and accounting for the resource usage of these Internet-scale applications presents a challenging technical problem. This paper ...
Fully decentralized emulation of best-effort and processor sharing queues
SIGMETRICS '08Control of large distributed cloud-based services is a challenging problem. The Distributed Rate Limiting (DRL) paradigm was recently proposed as a mechanism for tackling this problem. The heuristic nature of existing DRL solutions makes their behavior ...
Load balancing vs. distributed rate limiting: an unifying framework for cloud control
ICC'09: Proceedings of the 2009 IEEE international conference on CommunicationsWith the expansion of cloud-based services, the question as to how to control usage of such large distributed systems has become increasingly important. Load balancing (LB), and recently proposed distributed rate limiting (DRL) have been used ...
Comments