International Journal on Science and Technology

E-ISSN: 2229-7677     Impact Factor: 9.88

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Call for Paper Volume 16 Issue 2 April-June 2025 Submit your research before last 3 days of June to publish your research paper in the issue of April-June.

Achieving High Write Availability in Distributed Systems through Multi-Leader Replication

Author(s) Naveen Srikanth Pasupuleti
Country United States
Abstract In distributed systems, precise time synchronization is crucial for ensuring consistency and coordination between nodes, especially when performing tasks like logging, data replication, and fault detection. The Network Time Protocol (NTP) is one of the most widely used protocols for synchronizing clocks in computer networks. However, while NTP plays a critical role in time synchronization, it can suffer from performance issues, particularly when handling a large number of nodes or when the network experiences high latency. NTP synchronization time refers to the time taken by a node to synchronize its clock with an NTP server. These messages include timestamps that are used to calculate round-trip delays and adjust the system clock. In a large-scale system, this process can take longer, especially if the network latency is high or if the system relies on distant NTP servers that are geographically far from the nodes. This results in increased synchronization times, which may cause a delay in critical processes like data replication or the execution of time-sensitive tasks. Moreover, NTP’s reliance on a hierarchical time system, where each node communicates with a higher-level NTP server, can further contribute to synchronization delays. As the number of nodes increases, so does the complexity of the synchronization process, leading to increased overhead and slower synchronization times. In systems with a large number of nodes, the high NTP synchronization time can lead to inconsistencies in data replication and cause issues in maintaining strong consistency guarantees. These inconsistencies can cause problems such as race conditions, incorrect data, or system failures, particularly in distributed databases or applications that require a strong consistency model. As a result, it becomes necessary to explore ways to mitigate the high NTP synchronization times in large-scale distributed systems. Reducing synchronization time can improve the overall system performance, minimize inconsistencies, and ensure that time-sensitive tasks are executed more efficiently. Optimizing the synchronization process and utilizing faster time synchronization techniques or protocols can alleviate the performance issues associated with NTP synchronization in large clusters. By addressing these challenges, distributed systems can function more efficiently and consistently, providing better service and reliability. This paper addresses these issue by using Chrony Sync time.
Field Engineering
Published In Volume 13, Issue 2, April-June 2022
Published On 2022-05-12
DOI https://doi.org/10.71097/IJSAT.v13.i2.5841
Short DOI https://doi.org/g9p7mr

Share this