Peningkatan Kinerja Sistem Spatial Data Warehouse Kebakaran Hutan Menggunakan GeoServer dan GeoMondrian

Abstract

Forest fire can be periodically detected by satellite as hotspot areas that occur at specified time and location. The frequency of forest fire in Indonesia is high so the data collection of hotspots will be very large. That data collection of hotspots can be stored into specific storage such as data warehouse. A data warehouse system of forest fire was built in the previous research using GeoServer as web map server and GeoMondrian as OLAP Server. However, that system still has some limitations such as the poor system runtime and the low amount of retrieved hotspots. In this research, the system would be reconfigurated in order to fix those limitations so that the system performance would increase. The configuration involves some part of the system including Java Runtime Environment, database source, data cube schema, and error handling in OpenLayers. It has resulted 155 ms higher speed of runtime and 1310 higher number of retrieved hotspots.