Methods of Resource Allocation with Conflict Detection

The problems of resource allocation in query-applying pattern widely exist in daily applications. Resource allocation has the characteristic of exclusiveness. Lock methods are mostly used to deal with possible conflicts in the allocation process in traditional solutions. There is a natural contradiction between lock mechanism and multi task parallelism and such contradiction becomes more and more protruding with the increasing popularity of parallel processing environments represented by multi core CPU. A method of abandoning lock mechanism at application level is discussed in this paper. Instead, the method of resource conflict detection is used to ensure the correctness of transaction parallelism. According to sufficient conditions for the effectiveness of conflict detection, two kinds of conflict detection algorithms are proposed, which are separately based on the queue and on Key-T table. Then, the problem of resource allocation is further extended to "renting" situation with "time dimension", where a more general conflict detection algorithm is proposed. The parallel efficiency of multi task system can be improved remarkably and the implementation of the system can be much simplified with the scheme based on conflict detection.

[1]  Suresh Kumar,et al.  Secure concurrency control algorithm for multilevel secure databases , 2013 .

[2]  Hugo Rito,et al.  Adaptive transaction scheduling for mixed transactional workloads , 2015, Parallel Comput..

[3]  A.R. Ali,et al.  Two phase locking concurrency control in distributed database with N-tier architecture , 2004, International Conference on Electrical, Electronic and Computer Engineering, 2004. ICEEC '04..

[4]  Harumi A. Kuno,et al.  Concurrency Control for Adaptive Indexing , 2012, Proc. VLDB Endow..

[5]  Rajashree Shedge,et al.  Comparative Study of Concurrency Control Techniques in Distributed Databases , 2014, 2014 Fourth International Conference on Communication Systems and Network Technologies.

[6]  Stanley B. Zdonik,et al.  On Predictive Modeling for Optimizing Transaction Execution in Parallel OLTP Systems , 2011, Proc. VLDB Endow..

[7]  Jia Liu,et al.  A long-running transaction model of workflow , 2010, 2010 3rd International Conference on Biomedical Engineering and Informatics.

[8]  Teruhisa Hochin,et al.  Generation of Semantic Concurrency Control Program by Using Genetic Programming , 2012, 2012 IEEE/ACIS 11th International Conference on Computer and Information Science.

[9]  Waqar Haque,et al.  Dynamic Load Balancing in Real-Time Distributed Transaction Processing , 2013, 2013 IEEE 16th International Conference on Computational Science and Engineering.

[10]  Elyès Kooli,et al.  Predictive speculative concurrency control for real-time database systems , 2014, 2014 Information and Communication Technologies Innovation and Application (ICTIA).

[11]  Kirk W. Cameron,et al.  LUC: Limiting the Unintended Consequences of Power Scaling on Parallel Transaction-Oriented Workloads , 2015, 2015 IEEE International Parallel and Distributed Processing Symposium.

[12]  Grigorios Loukides,et al.  A Parallel Method for Scalable Anonymization of Transaction Data , 2015, 2015 14th International Symposium on Parallel and Distributed Computing.