Abstract
Caching systems using the Least Recently Used (LRU) principle have now become ubiquitous. A fundamental question for these systems is whether the cache space should be pooled together or divided to serve multiple flows of data item requests in order to minimize the miss probabilities. In this paper, we show that there is no straight yes or no answer to this question, depending on complex combinations of critical factors, including, e.g., request rates, overlapped data items across different request flows, data item popularities and their sizes. To this end, we characterize the performance of multiple flows of data item requests under resource pooling and separation for LRU caching when the cache size is large.
Analytically, we show that it is asymptotically optimal to jointly serve multiple flows if their data item sizes and popularity distributions are similar and their arrival rates do not differ significantly; the self-organizing property of LRU caching automatically optimizes the resource allocation among them asymptotically. Otherwise, separating these flows could be better, e.g., when data sizes vary significantly. We also quantify critical points beyond which resource pooling is better than separation for each of the flows when the overlapped data items exceed certain levels. Technically, for a broad class of heavy-tailed distributions we derive the asymptotic miss probabilities of multiple flows of requests with varying data item sizes in a shared LRU cache space. It also validates the characteristic time approximation under certain conditions. These results provide new insights on improving the performance of caching systems.
- Susanne Albers and Jeffery Westbrook. 1998. Self-organizing data structures. Online Algorithms: The state of the art Vol. 1442 (1998), 13--41. Google Scholar
Digital Library
- Brian Allen and Ian Munro. 1978. Self-organizing binary search trees. J. ACM Vol. 25 (1978), 526--535. Google Scholar
Digital Library
- Noga Alon and Joel H. Spencer. 2000. The probabilistic method (2nd ed.). John Wiley.Google Scholar
- Sigal Ar, Bernard Chazelle, and Ayellet Tal. 2000. Self-customized BSP trees for collision detection. Computational Geometry: Theory and Applications Vol. 10 (2000), 23--29. Google Scholar
Digital Library
- Martin Arlitt, Rich Friedrich, and Tai Jin. 1999. Workload characterization of a Web proxy in a cable modem environment. SIGMETRICS Performance Evalation Review Vol. 27, 2 (Sept. 1999), 25--36.. 2010. A fluid limit for a cache algorithm with general request processes. Advances in Applied Probability Vol. 42, 3 (2010), 816--833. Google Scholar
Digital Library
- Stefan Podlipnig and Laszlo Böszörmenyi. 2003. A survey of web cache replacement strategies. ACM Computing Surveys (CSUR) Vol. 35, 4 (Dec.. 2003), 374--398. ISSN0360-0300 Google Scholar
Digital Library
- Guocong Quan, Kaiyi Ji, and Jian Tan. 2018. LRU caching with dependent competing requests. IEEE INFOCOM 2018 - IEEE Conference on Computer Communications (INFOCOM 2018). Honolulu, USA.Google Scholar
Cross Ref
- Ronald Rivest. 1976. On self-organizing sequential search heuristics. Commun. ACM Vol. 19 (1976), 63--67. Google Scholar
Digital Library
- James Roberts and Nada Sbihi. 2013. Exploring the memory-bandwidth tradeoff in an information-centric network Teletraffic Congress (ITC), 2013 25th International. IEEE, 1--9.Google Scholar
- Liam Roditty and Uri Zwick. 2004. A fully dynamic reachability algorithm for directed graphs with an almost linear update time Proceedings of the 36th STOC. 184--191. Google Scholar
Digital Library
- Elisha J. Rosensweig, Jim Kurose, and Don Towsley. 2010. Approximate models for general cache networks. In Proceedings of the 29th Conference on Information Communications (INFOCOM'10). IEEE Press, San Diego, California, USA, 1100--1108. Google Scholar
Digital Library
- Albers S. and M. Mitzenmacher. 1998. Average case analyses of list update algorithms. Algorithmica Vol. 21 (1998), 312--329.Google Scholar
Cross Ref
- Daniel D. Sleator and Robert E. Tarjan. 1985 a. Amortized efficiency of list update and paging rules. Commun. ACM Vol. 28 (1985), 202--208. Google Scholar
Digital Library
- Daniel D. Sleator and Robert E. Tarjan. 1985 b. Self-adjusting binary search trees. J. ACM Vol. 32 (1985), 652--686. Google Scholar
Digital Library
- Alexander L. Stolyar. 2004. MaxWeight scheduling in a generalized switch: State space collapse and workload minimization in heavy traffic. The Annals of Applied Probability Vol. 14, 1 (02. 2004), 1--53.Google Scholar
Cross Ref
- Toyoaki Sugimoto and Naoto Miyoshi. 2006. On the asymptotics of fault probability in least-recently-used caching with Zipf-type request distribution. Random Structures & Algorithms Vol. 29, 3 (2006), 296--323. Google Scholar
Digital Library
- Jian Tan, Li Zhang, and Yandong Wang. 2015. Miss behavior for caching with lease. SIGMETRICS Performance Evaluation Review, MAMA workshop, Vol. 43, 2 (2015), 60--62. Google Scholar
Digital Library
- Andrew S. Tanenbaum. 2001. Modern Operating Systems (2rd ed.). Prentice Hall Press, Upper Saddle River, NJ, USA. Google Scholar
Digital Library
- Naoki Tsukada, Ryo Hirade, and Naoto Miyoshi. 2012. Fluid limit analysis of FIFO and RR caching for independent reference models. Performance Evaluation Vol. 69, 9 (2012), 403--412. Google Scholar
Digital Library
- Jia Wang. 1999. A survey of Web caching schemes for the Internet. SIGCOMM Computer Communication Review Vol. 29, 5 (Oct.. 1999), 36--46. ISSN 0146-4833 Google Scholar
Digital Library
- Xingbo Wu, Li Zhang, Yandong Wang, Yufei Ren, Michel Hack, and Song Jiang. 2016. zExpander: a key-value cache with both high performance and fewer misses Proceedings of the Eleventh European Conference on Computer Systems (EuroSys '16). ACM, New York, NY, USA, Article 14, 15 pages. Google Scholar
Digital Library
- Yuehai Xu, Eitan Frachtenberg, Song Jiang, and Mike Paleczny. 2014. Characterizing Facebook's Memcached workload. IEEE Internet Computing Vol. 18, 2 (2014), 41--49.Google Scholar
Cross Ref
- Yuehai Xua, Eitan Frachtenbergb, and Song Jiang. 2014. Building a high-performance key-value cache as an energy-efficient appliance. Performance Evaluation Vol. 79 (September. 2014), 24--37.Google Scholar
- Yue Yang and Jianwen Zhu. 2016. Write skew and Zipf distribution: Evidence and implications. ACM Transactions on Storage (TOS) Vol. 12, 4, Article 21 (June. 2016), pages 19 pages. ISSN1553-3077 Google Scholar
Digital Library
Index Terms
On Resource Pooling and Separation for LRU Caching
Recommendations
A New Flexible Multi-flow LRU Cache Management Paradigm for Minimizing Misses
The Least Recently Used (LRU) caching and its variants are used in large-scale data systems in order to provide high-speed data access for a wide class of applications. Nonetheless, a fundamental question still remains open: in order to minimize miss ...
On Resource Pooling and Separation for LRU Caching
SIGMETRICS '18Caching systems using the Least Recently Used (LRU) principle have now become ubiquitous. A fundamental question for these systems is whether the cache space should be pooled together or divided to serve multiple flows of data item requests in order to ...
On Resource Pooling and Separation for LRU Caching
SIGMETRICS '18: Abstracts of the 2018 ACM International Conference on Measurement and Modeling of Computer SystemsCaching systems using the Least Recently Used (LRU) principle have now become ubiquitous. A fundamental question for these systems is whether the cache space should be pooled together or divided to serve multiple flows of data item requests in order to ...






Comments