Towards a Highly Available Model for Processing Service Requests Based on Distributed Hash Tables

Voichiţa Iancu,Nicolae Ţăpuş

doi:10.3390/math10050831

Abstract

This work aims to identify techniques leading to a highly available request processing service by using the natural decentralization and the dispersion power of the hash function involved in a Distributed Hash Table (DHT). High availability is present mainly in systems that: scale well, are balanced and are fault tolerant. These are essential features of the Distributed Hash Tables (DHTs), which have been used mainly for storage purposes. The novelty of this paper’s approach is essentially based on hash functions and decentralized Distributed Hash Tables (DHTs), which lead to highly available data solutions, which a main building block to obtain an improved platform that offers high availability for processing clients’ requests. It is achieved by using a database constructed also on a DHT, which gives high availability to its data. Further, the model requires no changes in the interface, that the request processing service already has towards its clients. Subsequently, the DHT layer is added, for the service to run on top of it, and also a load balancing front end, in order to make it highly available, towards its clients. The paper shows, via experimental validation, the good qualities of the new request processing service, by arguing its improved scalability, load balancing and fault tolerance model.

Highlights

The concept of high availability represents a key feature for today’s online services
For the entire improved solution of a highly available service, we are most interested in not introducing too much overhead by adding this solution’s extra components, which yield high availability for the service. The elements that this novel solution aimed to improve, in order to make the service highly available, are: (1) static or dynamic load balancing, in order to maintain a high availability for the service, even in the context of increased load, due to a larger number of clients; (2) fault tolerance, by proving complete functionality even in the presence of failing nodes; (3) scalability, by Distributed Hash Table (DHT) self-extension described and evaluated in [41,42], in order to be able to cover a higher number of clients and the increased needs of persistence, that critical sensitive data have, in order to obtain a generic method for making a previously centralized service to be highly available, with minimum changes at the service level
This translates in the fact that there is a certain cost for the benefits of load balancing and decentralization, fault tolerance and scalability that Chord brings with it, but this time cost is perceived only for the first message of the session initiated by the client, and in cases of re-connection

Summary

Introduction

The concept of high availability represents a key feature for today’s online services. Encompasses multiple abstractization layers, data aggregation [5] and some peer-to-peer techniques, aspects the current work addresses too, in its specific novel manner oriented towards high availability through DHTs, as is described in what follows Nowadays, it is very important for any widely spread application service to be always able to answer and process correctly as many clients and client requests as it is needed, without having downtime. High availability is an important aspect concerning software applications, which is synonymous with the combination of: fault tolerance, scalability and load balancing among the system’s internal components, together with data coherence and data persistence Since most of these identified characteristics, to influence a system’s high availability, are very much related to not losing the data the system needs to process, and always obtaining the correct version of the data, this paper argues and stresses the fact that, a very good and highly available database to store the service’s data is a must for obtaining a highly available service. Available online: https://www.geeksforgeeks.org/introduction-of-virtual-routerredundancy-protocol-vrrp-and-its-configuration/ (accessed on 3 January 2022)

Objectives

Methods

Results

Discussion

Conclusion