An autonomic approach for fault tolerance using scaling, replication and monitoring in cloud computing

Abstract

The reason of grid monitoring and management is to detect offerings in grid surroundings for fault detection, performance evaluation, performance tuning, load balancing and scheduling. We have reviewed and worked on a self-managing fault-tolerance framework for high availability, scalability and reliability in cloud data center with proper monitoring tool which is deployed for Cluster monitoring and balancing purpose. This framework introduces the load balancers combined with Ganglia Monitoring System. The proposed framework is designed to overcome the issues regarding the high availability, fault tolerance, major downtime, self-healing, etc in cloud data center. HAProxy has been used to offer scaling to the servers for load balancing in proactive way. It additionally monitors the web servers for fault prevention on the client level. Our structure works with self-managed mirroring and load balancing in database servers with the help of MySQL master, replication of master database and Nginx consequently. Administrator supervise the running of servers through Ganglia Monitoring System because 24x7 manual monitoring cannot be possibly done by the cloud data center. For that we have carried out this scenario by using the virtualization technique in cloud environment.

A hybrid fault tolerance framework for SaaS services based on hidden Markov model

“…Owing to the complexity and heterogeneity (due to different types of techniques) of cloud computing solutions, there are many possible and novel causes or situations which may impact the dependability of the cloud services. Therefore, there are various fault tolerance solutions appeared in the research literatures (such as Bala and Chana, 2012;Amoon, 2015;Jhawar and Santambrogio, 2013;Garg and Bagga, 2015;Hong and Feuerlicht, 2016;Rosiello et al, 2014;Abraham et al, 2011 andCamarinha-Matos et al, 2015). In Bala and Chana (2012), techniques for creating fault tolerance in cloud computing are mentioned, and architectures proposed are introduced subsequently.…”