Enhancement of Generalized Earliest Deadline First Policy

被引:1
|
作者
Kaddes, Mourad [1 ,2 ]
Amanton, Laurent [1 ]
Berred, Alexandre [3 ]
Sadeg, Bruno [1 ]
Abdouli, Majed [2 ]
机构
[1] Univ Havre, LITIS, F-76600 Lehavre, France
[2] Sfax Univ, Pole Technol Sfax, MIR CL, Sfax 3021, Tunisia
[3] Univ Havre, LMAH, F-76600 Lehavre, France
关键词
Real-time Database System; Transaction Processing; Scheduling Policy; Success Ratio; Transaction Importance; TIME DATABASE-SYSTEMS; UPDATE TRANSACTIONS;
D O I
10.5220/0004448802310238
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Scheduling transactions in real-time database systems (RTDBSs) is more complex than scheduling tasks in real-time systems. In fact, the RTDBS must guarantee the database logical consistency, on one hand, and it must schedule the transactions in order to meet their deadlines, on the other hand. The main policy used to schedule transactions in RTDBSs is Earliest Deadline First (EDF). However, it is well-known that EDF is not efficient for scheduling transactions in overload conditions. Consequently, different scheduling protocols: AED, AEVD, APP, AEDF-Co, GEDF... were proposed to improve the system performances in firm RTDBSs, i.e. late transactions are considered useless. Generalized Earliest Deadline First (GEDF) is a new scheduling protocol in which transaction priority is assigned according to both deadlines and a parameter, called SPriority, which expresses the importance of transactions. In this paper, an RTDBSs analysis is presented. The accuracy of GEDF scheduling policy and the influence of database workload on the system performances is investigated. This study enabled us to describe the complete behavior of the transaction success ratio. Moreover, based on intensive simulations, we have derived the optimal values of the system parameters which improve the success ratio without modifying GEDF protocol
引用
收藏
页码:231 / 238
页数:8
相关论文
共 50 条
  • [31] Earliest deadline first scheduling technique for different networks in network control system
    C. T. Kalaivani
    N. Kalaiarasi
    Neural Computing and Applications, 2019, 31 : 223 - 232
  • [32] A probabilistic approach to estimate the mean waiting times in the earliest deadline first polling
    Chen, Mu-Song
    Yen, Hao-Wei
    COMPUTERS & INDUSTRIAL ENGINEERING, 2013, 64 (03) : 817 - 826
  • [33] A Simulation of Priority Based Earliest Deadline First Scheduling for Cloud Computing System
    Gupta, Garima
    Kumawat, Vimal Kr.
    Laxmi, P. R.
    Singh, Dharmendra
    Jain, Vinesh
    Singh, Ravinder
    2014 FIRST INTERNATIONAL CONFERENCE ON NETWORKS & SOFT COMPUTING (ICNSC), 2014, : 35 - 39
  • [34] Earliest deadline first scheduling technique for different networks in network control system
    Kalaivani, C. T.
    Kalaiarasi, N.
    NEURAL COMPUTING & APPLICATIONS, 2019, 31 (Suppl 1): : 223 - 232
  • [35] Scalable hardware earliest-deadline-first scheduler for ATM switching networks
    Kim, BK
    Shin, KG
    18TH IEEE REAL-TIME SYSTEMS SYMPOSIUM, PROCEEDINGS, 1997, : 210 - 218
  • [36] Asymptotic behavior of earliest-deadline-first scheduling with a large number of flows
    Han, YJ
    La, RJ
    Ren, TM
    2004 43RD IEEE CONFERENCE ON DECISION AND CONTROL (CDC), VOLS 1-5, 2004, : 3295 - 3300
  • [37] Minimizing cache usage with fixed-priority and earliest deadline first scheduling
    Sun, Binqi
    Kloda, Tomasz
    Garcia, Sergio Arribas
    Gracioli, Giovani
    Caccamo, Marco
    REAL-TIME SYSTEMS, 2024, : 625 - 664
  • [38] Statistical analysis of delay bound violations at an earliest deadline first (EDF) scheduler
    Sivaraman, V
    Chiussi, FM
    PERFORMANCE EVALUATION, 1999, 36-7 : 457 - 470
  • [39] Earliest Deadline First Scheduling for Real-Time Computing in Sustainable Sensors
    Chetto, Maryline
    El Osta, Rola
    SUSTAINABILITY, 2023, 15 (05)
  • [40] Earliest-deadline-first service in heavy-traffic acyclic networks
    Kruk, L
    Lehoczky, J
    Shreve, S
    Yeung, SN
    ANNALS OF APPLIED PROBABILITY, 2004, 14 (03): : 1306 - 1352