Zefektivnění monitoringu cloudové infrastruktury ve společnosti poskytující mailové kolaborativní řešení

The thesis deals with the improvement of cloud infrastructure monitoring in a company providing mail collaborative solution provided in the form of Software as a Service. The current implementation of monitoring includes various issues including separate monitoring of customer servers from other cloud infrastructure, long-term data storing in the same system and related performance problems of the Zabbix monitoring server. The aim of the thesis is to improve the monitoring of cloud infrastructure and to eliminate the deficiencies. Based on the analysis of the current state of infrastructure and its monitoring, selected issues are defined and tools are chosen to achieve the goal according to the company's requirements. Following is the implementation of monitoring in high availability mode, partial consolidation of individual elements and separation of long-term trends into a separate system. When implementing test environment, great emphasis is placed on setting the dependencies of individual elements to clarify and reveal the root cause of a potential problem. The result of the work is a Zabbix cluster in high availability mode, which extends the monitoring of customer servers to fundamental infrastructure components. The cluster is used exclusively for alerting with configured dependencies. Prometheus server and Grafana are created for working with historical data. The company can use the solution as a replacement for the current implementation.