A volcano monitoring system plays a key role for launching emergency early warning, and the use of alternative technologies have proven their effectiveness in this setting, which is the case of wireless sensor networks. For surveillance systems, the real-time requirement is mandatory due to the need for immediate access to the signals derived from a natural disaster where the goal is safeguarding lives. Previous works did not report detailed enough performance evaluation of this kind of systems, either by means of using simulation tools or in a test-bed related to real-time metrics. Our aim was to identify the optimum number of sensors to be deployed in a Volcano Monitoring System based on simulation results and corroborated with an in-situ testbed. We used ns-2 as simulation tool, where Random and Tessellation scenarios were evaluated. Our study identified that the optimal scenario in volcano monitoring is Random, with maximum eighteen nodes to satisfy metrics such as throughput, time delay, and packet loss. We deployed sixteen sensors in a strategic area at Cotopaxi Volcano, where the information was obtained during three days of continuous monitoring. This information was sent to a surveillance laboratory located 45 km away from the station placed at the volcano, and WiFi-based long distance technology was used for this purpose. The data obtained with our system allowed to distinguish long period events and volcano tectonic earthquakes.