No abstract

No abstract

The paper describes the universal approach for monitoring the data storage of a globally distributed cloud computing system, which allows you to automate creation of new metrics in the system and predict their behavior for the end users. Since the existing monitoring software products provide built-in scheme only for system metrics like RAM, CPU, disk drives, network traffic, but don’t offer solutions for business functions, IT companies have to design specialized database structure (DB). The data structure proposed in this paper for storing the monitoring statistics is universal and allows you to save resources when orginizing database monitoring on the scale of the GDCCS. The goal of the research is to develop a universal model for monitoring and forecasting of data storage in a globally distributed cloud computing system and its adequacy to real operating conditions.

Quality of IT Services (QoS), providing across all globally distributed regions via Internet, use modern cloud computing IT technologies having big data flow and, therefore, is actual. This paper briefly describes the methods of analysis and visualization of monitoring big data based on Key Performance Indicators (KPIs) of a cloud computing IT system using real world example of globally distributed infrastructure of the International IT Company. Implementation of proposed methods of visual analytics in worldwide leading IT companies – RingCentral (USA) and Zabbix (Latvia), – allowed improving of QoS and availability of IT services up to a worldwide level of 99.999% in 24/7 mode. Implementation of new solutions in IT companies is confirmed by corresponding documents and by publications in PhD and DSc thesis of the coauthors.