Virtual Differential Storage Based K-Rollback Concurrency Control Algorithm in Distributed Shared Memory Systems
SAbhinav Aggarwal1, Rupika Srivastava2, Sumit Malik3, Kirti Meena4, Poonam5
1Abhinav Aggarwal, Department of Computer Science and Engineering, Indian Institute of Technology (IIT) Roorkee, Roorkee, India.
2Rupika Srivastava, Department of Electronics and Computer Engineering, Indian Institute of Technology (IIT) Roorkee, Roorkee, India.
3Sumit Malik, Department of Computer Science and Engineering, Indian Institute of Technology (IIT) Roorkee, Roorkee, India.
4Kirti Meena, Department of Computer Science and Engineering, Indian Institute of Technology (IIT) Roorkee, Roorkee, India.
5Poonam, Department of Computer Science and Engineering, Indian Institute of Technology (IIT) Roorkee, Roorkee, India.
Manuscript received on August 05, 2013. | Revised Manuscript received on August 11, 2013. | Manuscript published on August 15, 2013. | PP: 52-58 | Volume-1 Issue-9, August 2013. | Retrieval Number: I0419081913/2013©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Most of the algorithms that exist today for concurrency control over distributed shared memory, either fail to provide a scalable model or involve a large communication overhead for establishing consensus over the state of the shared variables. After a thorough study of some of the efficient algorithms this field, this paper introduces a functional view of a holistic approach, which exploits the best features of all others. It provides a virtual differential storage, which allows fast replication and compact storage, along with a strong subversion control over rollbacks in time, which provides better fault tolerance. It also talks of an intelligent logging mechanism, where the read/write records are used actively by the central controller to provide exclusion over Above all, the algorithm is best implemented in LISP or Scheme due to its functional nature. This make the implementation computationally very fast. A trade off, however, exists between the implementation complexity and the quality of the final product.
Keywords: Log, Page, Concurrency, Shared Memory.