Scheduling algorithms for parallel transaction processing systems

  • Jiahong Wang
  • Jie Li
  • Hisao Kameda
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1277)


Shared-nothing parallel transaction processing (TP) systems have great potential to serve the ever-increasing demands for high transaction processing rate. This potential, however, may not be reached due to the negative effect of the widely used two-phase locking (2PL) concurrency control method. We observed that for the shared-nothing parallel TP system, this negative effect of 2PL can be alleviated significantly by scheduling transactions judiciously. In this paper, a new transaction scheduling algorithm called FCFSP (FCFS with Priority) is proposed thereby. In order to study the performance of transaction scheduling algorithms, a comprehensive simulator for shared-nothing parallel TP systems is developed. Using the developed simulator, the performance of FCFSP is compared with that of the conventional FCFS and the previously proposed SCST (Synchronizing Completion of SubTransactions) transaction scheduling algorithms. Simulation results demonstrate the effectiveness of FCFSP. Simulation results also show that FCFSP outperforms FCFS greatly, and overcomes the drawback of SCST.


Concurrency control parallel database performance evaluation simulation model transaction scheduling two-phase locking 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Bhide A.: An Analysis of Three Transaction Processing Architectures, Proc. 14th VLDB Conf., pp.339–350 (1988).Google Scholar
  2. 2.
    Boral, H., Alexander, W., Clay, L., Copeland, G., Danforth, S., Franklin, M., Hart, B., Smith, M. and Valduriez, P.: Prototyping Bubba, A Highly Parallel Database System, IEEE Trans. Knowledge Data Eng., Vol.2, No. 1, pp.4–24 (1990).Google Scholar
  3. 3.
    Carey, M.J., Krishnamurthy, S. and Livny, M.: Load Control for Locking: The ‘Half-and-Half’ Approach, Proc. 9th ACM PODS Conf., pp.72–84 (1990).Google Scholar
  4. 4.
    Dandamudi, S.P. and Chow, C.Y.: Performance of Transaction Scheduling Policies for Parallel Database Systems, Proc. 11th IEEE Conf. Distributed Computing Sys., pp.116–124 (1991).Google Scholar
  5. 5.
    DeWitt, D.J., Gerber, R., Graefe, G., Heytens, M., Kumar, K. and Muralikrishna, M.: GAMMA — A High Performance Dataflow Database Machine, Proc. 12th VLDB Conf., pp.25–28 (1986).Google Scholar
  6. 6.
    Franaszek, P.A., Haritsa, J.R., Robinson, IT. and Thomasian, A.: Distributed Concurrency Control Based on Limited Wait-Depth, IEEE Trans. Parallel Distributed Sys., Vol.4, No.11, pp.1246–1264 (1993).Google Scholar
  7. 7.
    Hsu, M. and Zhang, B.: Performance Evaluation of Cautious Waiting, ACM Trans. Database Sys., Vol.17, No.3, pp.477–512 (1992).Google Scholar
  8. 8.
    Jie Li, Jiahong Wang, Hisao Kameda, Keqin Li: An Effective Scheduling Algorithm for Parallel Transaction Processing Systems, to appear in Proc. the Special Session on Parallel and Distributed Computing Technology, IEEE National Aerospace and Electronics Conf., (Dayton, Ohio, USA, July 14–18, 1997).Google Scholar
  9. 9.
    Jenq, B.C., Twichell, B.C. and Keller, T.W.: Locking Performance in a Shared Nothing Parallel Database Machine, IEEE Trans. Knowledge Data Eng., Vol.1, No.4, pp.530–543 (1989).Google Scholar
  10. 10.
    Moenkeberg, A. and Weikum, G.: Conflict-Driven Load Control for the Avoidance of Data-Contention Thrashing, Proc. 7th IEEE Conf. Data Eng., pp.632–639 (1991).Google Scholar
  11. 11.
    Smith, M., Alexander, W., Boral, H., Copeland, G., Keller, T., Schwetman, H. and Young, C.R.: An Experiment on Response Time Scalability in Bubba, Proc. 6th Int. Workshop Database Machines, pp.35–57 (1989).Google Scholar
  12. 12.
    Stonebraker, M.: The Case for Shared Nothing, Database Eng. Bull., Vol.9, No.1, pp.4–9 (1986).Google Scholar
  13. 13.
    Tandem Database Group: NonStop SQL, A Distributed, High-Performance, High-Reliability Implementation of SQL, Proc. 2nd Workshop on High Performance Transaction Systems, pp.60–104 (1987).Google Scholar
  14. 14.
    Tay, Y.C., Goodman, N. and Suri, R.: Locking Performance in Centralized Databases, ACM Trans. Database Sys., Vol.10, No.4, pp.415–462 (1985).Google Scholar
  15. 15.
    Teradata: DBC/1012 Data Base Computer Concepts & Facilities, Document No. C02-001-00, Teradata Corp. (1983).Google Scholar
  16. 16.
    Thomasian, A.: Centralized Concurrency Control Methods for High-End TP, ACM SIGMOD Rec., Vol.20, No.3, pp.106–115 (1991).Google Scholar
  17. 17.
    Thomasian, A.: Two-Phase Locking Performance and Its Thrashing Behavior, ACM Trans. Database Sys., Vol.18, No.4, pp.579–625 (1993).Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • Jiahong Wang
    • 1
  • Jie Li
    • 1
  • Hisao Kameda
    • 1
  1. 1.Institute of Information Sciences and ElectronicsUniversity of TsukubaTsukuba Science City, IbarakiJapan

Personalised recommendations