Qiaomin Xie
Abstract
We consider a system of $N$ parallel servers, where each server consists of B units of a resource. Jobs arrive at this system according to a Poisson process, and each job stays in the system for an exponentially distributed amount of time. Each job may request different units of the resource from the system. The goal is to understand how to route arriving jobs to the servers to minimize the probability that an arriving job does not find the required amount of resource at the server, i.e., the goal is to minimize blocking probability. The motivation for this problem arises from the design of cloud computing systems in which the jobs are virtual machines (VMs) that request resources such as memory from a large pool of servers. In this paper, we consider power-of-d-choices routing, where a job is routed to the server with the largest amount of available resource among d ≥ 2 randomly chosen servers. We consider a fluid model that corresponds to the limit as N goes to infinity and provide an explicit upper bound for the equilibrium blocking probability. We show that the upper bound exhibits different behavior as B goes to infinity depending on the relationship between the total traffic intensity λ and B. In particular, if (B -- λ)/√λ → α, the upper bound is doubly exponential in √λ and if (B -- λ)/logd λ → β, β > 1, the upper bound is exponential in λ. Simulation results show that the blocking probability, even for small B, exhibits qualitatively different behavior in the two traffic regimes. This is in contrast with the result for random routing, where the blocking probability scales as O(1/√λ) even if (B -- λ)/√λ → α.
- Amazon EC2. http://aws.amazon.com/ec2/.Google Scholar
- V. Anantharam and M. Benchekroun. A technique for computing sojourn times in large networks of interacting queues. Probability in the Engineering and Informational Sciences, 7(04):441--464, 1993.Google ScholarCross Ref
- Google App Engine. https://cloud.google.com/appengine/docs?csw=1.Google Scholar
- Y. Azar, A. Z. Broder, A. R. Karlin, and E. Upfal. Balanced allocations. SIAM J. Comput., 29(1):180--200, Sept. 1999. Google ScholarDigital Library
- Azure. http://azure.microsoft.com/en-us/.Google Scholar
- N. Bansal, A. Caprara, and M. Sviridenko. A new approximation method for set covering problems, with applications to multidimensional bin packing. SIAM Journal on Computing, 39(4):1256--1278, 2010. Google ScholarDigital Library
- A. A. Borovkov. Stochastic Processes in Queueing Theory. Springer, 1976.Google ScholarCross Ref
- M. Bramson, Y. Lu, and B. Prabhakar. Asymptotic independence of queues under randomized load balancing. Queueing Systems, 71(3):247--292, 2012. Google ScholarDigital Library
- J. Csirik, D. S. Johnson, C. Kenyon, J. B. Orlin, P. W. Shor, and R. R. Weber. On the sum-of-squares algorithm for bin packing. J. ACM, 53(1):1--65, Jan. 2006. Google ScholarDigital Library
- C. Graham. Chaoticity on path space for a queueing network with selection of the shortest queue among several. Journal of Applied Probability, 37(1):198--211, 2000.Google ScholarCross Ref
- V. Gupta and A. Radovanovic. Online stochastic bin packing, 2012.Google Scholar
- J. Kaufman. Blocking in a shared resource environment. Communications, IEEE Transactions on, 29(10):1474--1481, Oct 1981.Google ScholarCross Ref
- T. G. Kurtz. Approximation of Population Processes. Society for Industrial and Applied Mathematics, 1981.Google ScholarCross Ref
- M. Luczak and C. McDiarmid. On the maximum queue length in the supermarket model. The Annals of Probability, 34(2):493--527, 2006.Google ScholarCross Ref
- S. Maguluri, R. Srikant, and L. Ying. Stochastic models of load balancing and scheduling in cloud computing clusters. In Proc. of IEEE INFOCOM, pages 702--710, Mar 2012.Google ScholarCross Ref
- S. T. Maguluri, R. Srikant, and L. Ying. Heavy traffic optimal resource allocation algorithms for cloud computing clusters. In Proc. of the 24th International Teletraffic Congress, pages 25:1--25:8, 2012. Google ScholarDigital Library
- X. Meng, V. Pappas, and L. Zhang. Improving the scalability of data center networks with traffic-aware virtual machine placement. In Proc. of IEEE INFOCOM, pages 1154--1162, Piscataway, NJ, USA, 2010. Google ScholarDigital Library
- M. Mitzenmacher. The power of two choices in randomized load balancing. PhD thesis, UC Berkeley, 1996. Google ScholarDigital Library
- M. Mitzenmacher. Studying balanced allocations with differential equations. Combinatorics, Probability and Computing, 8(5):473--482, Sept. 1999. Google ScholarDigital Library
- A. Mukhopadhyay and R. R. Mazumdar. Analysis of load balancing in large heterogeneous processor sharing systems. ArXiv preprint arXiv:1311.5806, 2013.Google Scholar
- A. Mukhopadhyay, R. R. Mazumdar, and F. Guillemin. Static versus dynamic user assignment to cloud resources. 2015.Google Scholar
- J. W. Roberts. A service system with heterogeneous user requirement. In G. Pujolle, editor, Performance of Data Communications Systems and Their Applications, 1981.Google Scholar
- R. Srikant and W. Whitt. Simulation run lengths to estimate blocking probabilities. ACM Trans. Model. Comput. Simul., 6(1):7--52, Jan. 1996. Google ScholarDigital Library
- A. L. Stolyar. An infinite server system with general packing constraints. ArXiv preprint arXiv:1205.4271, 2012.Google Scholar
- A. L. Stolyar and Y. Zhong. An infinite server system with general packing constraints: Asymptotic optimality of a greedy randomized algorithm. In Proc. 53th Annu. Allerton Conf. Commun., Control Comput., pages 575--582, Oct 2013.Google ScholarCross Ref
- A. L. Stolyar and Y. Zhong. A large-scale service system with packing constraints: Minimizing the number of occupied servers. SIGMETRICS Perform. Eval. Rev., 41(1):41--52, June 2013. Google ScholarDigital Library
- J. N. Tsitsiklis and K. Xu. On the power of (even a little) resource pooling. Stochastic Systems, 2(1):1--66, 2012.Google ScholarCross Ref
- N. D. Vvedenskaya, R. L. Dobrushin, and F. I. Karpelevich. Queueing system with selection of the shortest of two queues: An asymptotic approach. Probl. Peredachi Inf., 32(1):20--34, 1996.Google Scholar
- L. Wang, F. Zhang, A. V. Vasilakos, C. Hou, and Z. Liu. Joint virtual machine assignment and traffic engineering for green data center networks. SIGMETRICS Perform. Eval. Rev., 41(3):107--112, Jan. 2014. Google ScholarDigital Library
- W. Whitt. Heavy-traffic approximations for service systems with blocking. AT&T Bell Laboratories Technical Journal, 63(5):689--708, 1984.Google ScholarCross Ref
- L. Ying, R. Srikant, and X. Kang. The power of slightly more than one sample in randomized load balancing. In Proc. of IEEE INFOCOM, 2015.Google ScholarCross Ref
Index Terms
- Power of d Choices for Large-Scale Bin Packing: A Loss Model
Recommendations
Power of d Choices for Large-Scale Bin Packing: A Loss Model
SIGMETRICS '15: Proceedings of the 2015 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer SystemsWe consider a system of $N$ parallel servers, where each server consists of B units of a resource. Jobs arrive at this system according to a Poisson process, and each job stays in the system for an exponentially distributed amount of time. Each job may ...
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, ...
Call-burst blocking of ON-OFF traffic sources with retrials under the complete sharing policy
In this paper we calculate both call and burst blocking probabilities of ON-OFF traffic sources with retrials. Calls of service-classes arrive to a single link according to a Poisson process and compete for the available link bandwidth under the ...
Comments