CN Monitor: Community Network Monitoring Platform Using Big Data, Log Files and SNMP Protocol

The Internet is nowadays an almost indispensable tool to carry out our activities and projects. It has become a necessity, even a fundamental right to have access to the Internet. However, several regions of the globe, mainly in Africa, are still not connected to the Internet prompting governments to find solutions to solve this problem. One approach that has proven to be inexpensive for connecting the unconnected is the creation of community networks. These are wireless computer networks that are made available to people and allow them to access the Internet and other services. Although they are economically viable, community networks are still subject to numerous breakdowns. Indeed, this type of network, particularly the one in Bibémi that we are studying, is subject to failures in both the equipment and the network applications. Due to their location, which is generally difficult to access, as they are located in rural areas, a breakdown can last several days before being resolved. It is therefore important to be able to anticipate these failures in order to avoid possible breakdowns or to limit network downtime. One approach to achieve this would be the development of tools to supervise this type of network. Thus, as part of this work, we have developed a monitoring platform that uses the SNMP protocol, log files and the Big Data Spark structured streaming framework to know the operating parameters of equipment and applications of the community network of Bibémi and thus alert in case of failure. The SNMP protocol and the log files are used here as our source of data from the devices and applications respectively. To perform real-time processing, we used the Spark structured streaming engine to facilitate queries on the data set. In addition, the data we exploit are those of a network simulated with the GNS3 tool respecting thearchitecture of the Bibémi community network. Finally, a web application has been developed to have a global view of the network and to allow immediate alert in case of failure.