Abstract
We introduce a new framework for the analysis of large-scale load balancing networks with general service time distributions, motivated by applications in server farms, distributed memory machines, cloud computing and communication systems. For a parallel server network using the so-called $SQ(d)$ load balancing routing policy, we use a novel representation for the state of the system and identify its fluid limit, when the number of servers goes to infinity and the arrival rate per server tends to a constant. The fluid limit is characterized as the unique solution to a countable system of coupled partial differential equations (PDE), which serve to approximate transient Quality of Service parameters such as the expected virtual waiting time and queue length distribution. In the special case when the service time distribution is exponential, our method recovers the well-known ordinary differential equation characterization of the fluid limit.
Furthermore, we develop a numerical scheme to solve the PDE, and demonstrate the efficacy of the PDE approximation by comparing it with Monte Carlo simulations. We also illustrate how the PDE can be used to gain insight into the performance of large networks in practical scenarios by analyzing relaxation times in a backlogged network. In particular, our numerical approximation of the PDE uncovers two interesting properties of relaxation times under the SQ(2) algorithm. Firstly, when the service time distribution is Pareto with unit mean, the relaxation time decreases as the tail becomes heavier. This is a priori counterintuitive given that for the Pareto distribution, heavier tails have been shown to lead to worse tail behavior in equilibrium. Secondly, for unit mean light-tailed service distributions such as the Weibull and lognormal, the relaxation time decreases as the variance increases. This is in contrast to the behavior observed under random routing, where the relaxation time increases with increase in variance.
- Aghajani, R. and Ramanan, K. (2017). Hydrodynamic limit of a randomized load balancing network. arXiv:1707.02005 {math.PR}.Google Scholar
- Asmussen, S. (2003). Applied Probability and Queues. Springer-Verlag, 2nd edition edition.Google Scholar
- Azar, Y., Broder, A. Z., Karlin, A. R., and Upfal, E. (1999). Balanced allocations. SIAM J. Comput., 29(1):180--200. Google Scholar
Digital Library
- Billingsley, P. (1968). Convergence of Probability Measures. John Wiley, New York.Google Scholar
- Bramson, M., Lu, Y., and Prabhakar, B. (2010). Randomized load balancing with general service time distributions. SIGMETRICS Perform. Eval. Rev., 38(1):275--286. Google Scholar
Digital Library
- Bramson, M., Lu, Y., and Prabhakar, B. (2012). Asymptotic independence of queues under randomized load balancing. Queueing Systems, 71(3):247--292. Google Scholar
Digital Library
- Bramson, M., Lu, Y., and Prabhakar, B. (2013). Decay of tails at equilibrium for FIFO join the shortest queue networks. The Annals of Applied Probability, 23(5):1841--1878.Google Scholar
Cross Ref
- Brown, L., Gans, N., Mandelbaum, A., Sakov, A., Shen, H., Zeltyn, S., and Zhao, L. (2005). Statistical analysis of a telephone call center. Journal of the American Statistical Association, 100(469):36--50.Google Scholar
Cross Ref
- Chen, S., Sun, Y., Kozat, U., Huang, L., Sinha, P., Liang, G., Liu, X., and Shroff, N. (2014). When queueing meets coding: Optimal-latency data retrieving scheme in storage clouds. In INFOCOM, 2014 Proceedings IEEE, pages 1042--1050.Google Scholar
Cross Ref
- Courant, R., Isaacson, E., and Rees, M. (1952). On the solution of nonlinear hyperbolic differential equations by finite differences. Communications on Pure and Applied Mathematics, 5(3):243--255.Google Scholar
Cross Ref
- Dai, J., Dieker, A., and Gao, X. (2014). Validity of heavy-traffic steady-state approximations in many-server queues with abandonment. Queueing Systems, 78(1):1--29. Google Scholar
Digital Library
- Ethier, S. and Kurtz, T. (1986). Markov processes: characterization and convergence. Wiley series in probability and mathematical statistics. Probability and mathematical statistics. Wiley.Google Scholar
- Evans, L. (1998). Partial Differential Equations. Graduate Studies in Mathematics. American Math. Soc.Google Scholar
- Farias, V., Moallemi, C., and Prabhakar, B. (2005). Load balancing with migration penalties. In Information Theory, 2005. ISIT 2005. Proceedings. International Symposium on, pages 558--562.Google Scholar
Cross Ref
- Graham, C. (2000). Chaoticity on path space for a queueing network with selection of the shortest queue among several. Journal of Applied Probability, 37(1):198--211.Google Scholar
Cross Ref
- Grossmann, C., Roos, H., and Stynes, M. (2007). Numerical Treatment of Partial Differential Equations. Universitext. Springer Berlin Heidelberg.Google Scholar
- Kardassakis, K. (2014). Load balancing in stochastic networks: Algorithms, analysis, and game theory. Undergraduate Honors Thesis, Brown University.Google Scholar
- Kolesar, P. (1984). Stalking the endangered cat: A queueing analysis of congestion at automatic teller machines. Interfaces, 14(6):16--26. Google Scholar
Digital Library
- Liang, G. and Kozat, U. (2014). TOFEC: achieving optimal throughput-delay trade-off of cloud storage using erasure codes. In INFOCOM, 2014 Proceedings IEEE, pages 826--834.Google Scholar
Cross Ref
- Luczak, M. J. and McDiarmid, C. (2006). On the maximum queue length in the supermarket model. The Annals of Probability, 34(2):493--527.Google Scholar
Cross Ref
- Luczak, M. J. and Norris, J. (2005). Strong approximation for the supermarket model. The Annals of Applied Probability, 15(3):2038--2061.Google Scholar
Cross Ref
- Mitzenmacher, M. (2000). Analyses of load stealing models based on families of differential equations. Theory of Computing Systems, 34(1):77--98.Google Scholar
Cross Ref
- Mitzenmacher, M. (2001). The power of two choices in randomized load balancing. IEEE Trans. Parallel Distrib. Syst., 12(10):1094--1104. Google Scholar
Digital Library
- Mukherjee, D., Borst, S., van Leeuwaarden, J., and Whiting, P. (2016). Universality of power-of-d load balancing in many-server systems. arXiv:1612.00723 {math.PR}.Google Scholar
- Seelen, L. (1986). An algorithm for ph/ph/c queues. European Journal of Operational Research, 23(1):118 -- 127.Google Scholar
Cross Ref
- Vvedenskaya, N. D., Dobrushin, R. L., and Karpelevich, F. I. (1996). A queueing system with a choice of the shorter of two queues--an asymptotic approach. Problemy Peredachi Informatsii, 32(1):20--34.Google Scholar
Index Terms
The PDE Method for the Analysis of Randomized Load Balancing Networks
Recommendations
The PDE Method for the Analysis of Randomized Load Balancing Networks
SIGMETRICS '18We introduce a new framework for the analysis of large-scale load balancing networks with general service time distributions, motivated by applications in server farms, distributed memory machines, cloud computing and communication systems. For a ...
The PDE Method for the Analysis of Randomized Load Balancing Networks
SIGMETRICS '18: Abstracts of the 2018 ACM International Conference on Measurement and Modeling of Computer SystemsWe introduce a new framework for the analysis of large-scale load balancing networks with general service time distributions, motivated by applications in server farms, distributed memory machines, cloud computing and communication systems. For a ...
Randomized load balancing with general service time distributions
SIGMETRICS '10: Proceedings of the ACM SIGMETRICS international conference on Measurement and modeling of computer systemsRandomized load balancing greatly improves the sharing of resources in a number of applications while being simple to implement. One model that has been extensively used to study randomized load balancing schemes is the supermarket model. In this model, ...






Comments