Downloads: 1 | Views: 67 | Weekly Hits: ⮙1 | Monthly Hits: ⮙1
Research Paper | Computer Technology | India | Volume 7 Issue 11, November 2018 | Popularity: 5.2 / 10
Techniques for Managing Inconsistency to Scale Services in Distributed Memory Systems
Gnana Teja
Abstract: Distributed memory systems are essential commodities in present - day computation since they facilitate intricate computations in nodes that are often interconnected. However, as the masses of these systems increase, consistency management becomes increasingly more complex because of factors such as delays in the network, node crashes, and concurrent changes to the data. The scalability of various techniques of managing inconsistency in distributed memory systems is the concern of this paper. In the case of consistency models, this paper focuses on the aspects of strong and weak and their advantages and disadvantages. The strong consistency guarantee makes the data identical across multiple nodes, but this reduces the system's scalability and performance; hence, it is inefficient for applications like real - time data processing. Weak consistency models, in contrast, are less strict and allow some temporary data inconsistency, making them even more performant and scalable. However, they may need to provide more accurate data to consumers. The paper also focuses on the kinds of data, single primary users and multiple users, and how these determine the consistency models. Moreover, methods of consistency management of services evolving further are analyzed, as well as features and tendencies in the implementation of consistency management, including the hybrid and adaptive models and their perspective for growing services in the future. These models claim to balance how much of the actual data is captured. At the same time, it improves the system's response time, making it imperative, especially for developers and businesses aspiring to tackle massive distributed systems. The paper concludes and highlights that as distributed systems evolve to unprecedented complexity, scalability, and resilience of consistency will prove central to confidence in system integrity. It lays down the sum and substance of the existing methods and prospects that will be useful for the study's practitioner and theoretician.
Keywords: Distributed Memory Systems, Consistency Models, Scalability, Strong Consistency, Weak Consistency, Eventual, Consistency, Concurrency, Network Latency, Data Replication, Hybrid Consistency Models
Edition: Volume 7 Issue 11, November 2018
Pages: 2000 - 2008
Make Sure to Disable the Pop-Up Blocker of Web Browser