Research And Implementation Of A Monitoring System For A SaaS Platform

As SaaS (software as a service) model is more and more mature, the tenant for the use of cloud computing service is more and more widely, the quality of cloud computing service gradually get the attention of people. The data of tenant store in the cloud, the cloud errors, such as loss of data and system crash will bring great trouble to the tenant for using cloud services, and the reliability of the cloud computing platform is the foundation of the cloud service which tenants use. In order to improve the reliability of the cloud computing platform and the quality of services in cloud computing platform, it is necessary to build a monitoring system for the SaaS platform and the services on it so that we can get the real-time running status of every detail, also, the system will issue a warning notification when an error occurs. The goal of the monitoring system is to monitor the hardware resource occupied by the servers of the SaaS platform and the status of the services tenants use, and implement the storage and display of monitoring data. In addition, for multi-tenant SaaS platform applications, it is also necessary to analyze the data to get each tenant’s occupancy of CPU resource.The system is based on Nagios core, using the active and passive monitoring method, design to complete the whole architecture of distributed monitoring system, make it adapt to the multi-tenant cloud computing environment monitoring; design and implement the monitoring data acquisition module, monitoring data analysis and storage module, alarm module and the monitoring configuration module, and get each tenant occupancy of CPU resources from the monitoring data. The primary work of the paper is:1) Research on SaaS platform monitoring system, mainly research the related papers and open source tools, open source framework, extracting summarizes the requirements of the monitoring system for a SaaS platform.2) Analysis and design the architecture and function module of the monitoring system for a SaaS platform. Develop and customize the monitoring system based on Nagios core, which contains the distributed monitoring system design, monitoring data acquisition, storage and display of the data, and a configurable module of monitoring service.3) Gives a model of tenant CPU resource occupancy based on kalman filter, also implement a tenant occupancy program based on java bytecode transform, compare advantages and disadvantages of two schemes under the experimental conditions.