Performance Analysis of Key-Value Stores with Consistent Replica Selection Approach

  • Thazin NweEmail author
  • Tin Tin YeeEmail author
  • Ei Chaw HtoonEmail author
Conference paper
Part of the Communications in Computer and Information Science book series (CCIS, volume 1178)


A key-value store is the primary architecture of data centers. Most modern data stores tend to be distributed and to enable the scaling of the replicas and data across multiple instances of commodity hardware. Defining static replica placement mechanisms in different data centers lack the efficiency of the storage system. In the proposed system, dynamic scaling that changes the key/value store with replicas dynamically joining or leaving. To enhance the dynamic scaling of the replicas, the consistent hashing mechanism is enhanced in key-value stores due to the adaptability of node changes. This mechanism performs the eventual consistency services that offer quorum key-value store with increased consistency. According to the ordering of the hash values among the replicas in the ring, it could provide higher system throughput and reduce lower latency cost without using the random of the original consistent hashing method. An experimental result overwhelms the loss of original consistent hashing algorithms entirely and is proper for the distributed key-value store.


Key-value store Dynamic scaling Consistent hashing Eventual consistency Quorum 



I would also like to express my special thanks to Professor Junya Nakamura, Information and Media Center, the Toyohashi University, Japan for his valuable suggestions and guidelines.


  1. 1.
    Cooper, B.F., Silberstein, A., Tam, E., Ramakrishnan, R., Sears, R.: Benchmarking cloud serving systems with YCSB. In: Proceedings of the 1st ACM Symposium on Cloud computing, New York, pp. 143–154 (2010)Google Scholar
  2. 2.
    Malkhi, D., Reiter, M.K.: Byzantine quorum systems. In: Proceedings of the 29th Annual ACM Symposium on the Theory of Computing, STOC 1997, pp. 569–578, May 1997 Google Scholar
  3. 3.
    Karger, D., Lehman, E., Leighton, T., Panigraphy, R., Levine, M., Lewin, D.: Consistent hashing and random trees: distributed caching protocols for relieving hot spots on the world wide web. STOC 1997 Proceedings of the Twenty-Ninth Annual ACM Symposium on Theory of Computing, pp. 654–663 (1997)Google Scholar
  4. 4.
    Diogo, M., Cabral, B., Bernardino, J.: Consistency models of NoSQL databases. Future Internet 11 (2019). Scholar
  5. 5.
    Bailis, P., Venkataraman, S., Franklin, M.J., Hellerstein, J.M., Stoica, I.: Probabilistically bounded staleness for practical partial quorums. J. Proc. VLDB Endowment 5(8), 776–787 (2012). Scholar
  6. 6.
    Nwe, T., Nakamura, J., Yee, T.T., Phyu, M.P., Htoon, E.C.: Automatic adjustment of read consistency level of distributed key-value storage by a replica selection approach. In: The 1st International Conference on Advanced Information Technologies, Yangon, Myanmar, pp. 151–156, November 2017Google Scholar
  7. 7.
    Nwe, T., Yee, T.T., Htoon, E.C.: Improving read/write performance for key-value storage system by automatic adjustment of consistency level. In: The 33rd International Technical Conference on Circuits/Systems, Computers and Communications, Bangkok, Thailand, pp. 357–360, July 2018Google Scholar
  8. 8.
    Nwe, T., Nakamura, J., Yee, T.T., Htoon, E.C.: Automatic adjustment of consistency level by predicting staleness rate for distributed key-value storage system. In: Proceedings of the 2nd International Conference on Advanced Information Technologies, Yangon, Myanmar, pp. 27–32, November 2018Google Scholar
  9. 9.
    Nwe, T., Nakamura, J., Yee, T.T., Htoon, E.C.: A consistent replica selection approach for distributed key-value storage. In: Proceedings of 2019 International Conference on Advanced Information Technologies, Yangon, Myanmar, pp. 114–119, November 2019Google Scholar
  10. 10.
    Rahman, M.R., Golab, W., AuYoung, A., Keeton, K., Wylie, J.J.: Toward a principled framework for benchmarking consistency. In: Proceedings of the Eighth USENIX Conference on Hot Topics in System Dependability, ser. HotDep 2012. USENIX Association, Berkeley, p. 8 (2012)Google Scholar
  11. 11.
    Lakshman, A., Malik, P.: Cassandra: a decentralized structured storage system. SIGOPS Oper. Syst. Rev. 44(2), 35–40 (2010)CrossRefGoogle Scholar
  12. 12.
    DeCandia, G., et al.: Dynamo: amazon’s highly available keyvalue store. In: ACM SIGOPS Operating Systems Review, vol. 41, pp. 205–220. ACM (2007)Google Scholar
  13. 13.
    Nightingale, E.B., Elson, J., Fan, J., Hofmann, O.S., Howell, J., Suzue, Y.: Flat datacenter storage. In: OSDI, pp. 1–15 (2012)Google Scholar
  14. 14.
    Suresh, L., Canini, M., Schmid, S., Feldmann, A.: C3: cutting tail latency in cloud data stores via adaptive replica selection. In: 12th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2015, 4–6 May 2015, Oakland, CA, USA (2015)Google Scholar
  15. 15.
    Jaiman, V., Mokhtar, S.B., Quema, V., Chen, L.Y., Riviere, E.: He’ron: taming tail latencies in key-value stores under heterogeneous workloads. In: 2018 IEEE 37th Symposium on Reliable Distributed Systems (SRDS), pp. 191–200 (2018)Google Scholar
  16. 16.
    Dean, J., Barroso, L.A.: The tail at scale. Commun. ACM 56(2), 74–80 (2013)CrossRefGoogle Scholar
  17. 17.
    Bengfort, B., Xirogiannopoulos, K., Keleher, P.: Anti-entropy bandits for geo-replicated consistency. In: 2018 IEEE 38th International Conference on Distributed Computing Systems (2018)Google Scholar
  18. 18.
    Walters, J.P., Chaudhary, V.: Replication-based fault tolerance for MPI applications. IEEE Trans. Parallel Distrib. Syst. 20(7), 997–1010 (2009)CrossRefGoogle Scholar
  19. 19.
    Davoudian, A., Chen, L., Liu, M.: A survey on NoSQL stores. ACM Comput. Surv. (CSUR) 51, 1–43 (2018)CrossRefGoogle Scholar
  20. 20.
    Williams, B.: Dynamic snitching in Cassandra: past, present, and future (2012).

Copyright information

© Springer Nature Singapore Pte Ltd. 2020

Authors and Affiliations

  1. 1.University of Information TechnologyYangonMyanmar

Personalised recommendations