Time Series Based Replication Lag Modelling for Disaster Recovery in IT Sector

With the emerging IT infrastructures, disaster recovery practice is one of the indispensable parts. The systems should be maintained, and the data should be taken care of such that it will result in proper integrity of data and resilience of the system. In this paper, we propose a method based on time series analysis towards the very accurate computation of replication lag, one of the key metrics of disaster recovery systems. In other words, our approach uses time series data in one of the key stable platforms to forecast and monitor the replication lag between the primary and backup data centres. Our method will use statistical approaches like moving averages and trend analysis to enlighten the performance of replication processes. This will further bring proactive steps in mitigation of any probable data loss in disaster events from getting replicated incompletely. The significance of this research is an added value to the range of strategies of disaster recovery by implementing an approach to measure and manage replication lag to enhance the reliability and effectiveness of the disaster recovery system.