Reference Hub

This research has been cited in:

Chapter
Performance Issues in Scheduling of Real-Time TransactionsDatabase Systems for Advanced Applications10.1007/978-3-030-73200-4_51
Article
On Developing Framework for Schedulable Priority-Driven Systems: A Futuristic ReviewWireless Personal Communications10.1007/s11277-022-10082-9
Article
STEP: A Concomitant Protocol for Real Time ApplicationsWireless Personal Communications10.1007/s11277-021-09112-9

A Multi-Replica-Centered Commit Protocol for Distributed Real-Time and Embedded Applications

Anupama Arun, Sarvesh Pandey, Udai Shanker

Source Title: International Journal of System Dynamics Applications (IJSDA)10(4)

ISSN: 2160-9772|EISSN: 2160-9799|EISBN13: 9781799858997|DOI: 10.4018/IJSDA.20211001.oa18

Cite Article Cite Article

MLA

Arun, Anupama, et al. "A Multi-Replica-Centered Commit Protocol for Distributed Real-Time and Embedded Applications." IJSDA vol.10, no.4 2021: pp.1-19. http://doi.org/10.4018/IJSDA.20211001.oa18

APA

Arun, A., Pandey, S., & Shanker, U. (2021). A Multi-Replica-Centered Commit Protocol for Distributed Real-Time and Embedded Applications. International Journal of System Dynamics Applications (IJSDA), 10(4), 1-19. http://doi.org/10.4018/IJSDA.20211001.oa18

Chicago

Arun, Anupama, Sarvesh Pandey, and Udai Shanker. "A Multi-Replica-Centered Commit Protocol for Distributed Real-Time and Embedded Applications," International Journal of System Dynamics Applications (IJSDA) 10, no.4: 1-19. http://doi.org/10.4018/IJSDA.20211001.oa18

Export Reference

Favorite Full-Issue Download

View Full Text HTML

View Full Text PDF

Abstract

Modern multi-site database applications are not only time-driven but also require efficient quality of services with no single-node failure. It might be ideally achieved using database replication techniques. The transactions, being a basic component of these applications, are more likely to miss their deadlines because of requiring unpredictably long time to access remote data items. The temporal validity of data is another issue requiring attention to be paid. To address these problems, a Cluster-Replicas with Efficient Distributed Lazy Update (CRED) protocol is proposed in this paper. The CRED protocol increases the chance of timely execution of transactions and data freshness in an unpredictable workload environment by utilizing the lazy replica update strategy. It reduces the negative impact of the burst workload with a marginal overhead of ensuring timely-updated replicas. The simulation results confirm that the CRED outperforms the ORDER protocol by up to 4%.