Expert system in detection and correction of Grid nodes’ incidences
Document typeMaster thesis
Rights accessRestricted access - author's decision
The possibility of connecting huge range of computers sharing information, computing data and storage capacity is known as Grid Technology. Grid Technology or Distributed Computing is a network infrastructure that permits managing the available whole world computing power to process advanced computing applications. The main objective of this final project is to make progress in the automation of Grid operations, improving and extending an expert system designed to control a Grid node. InGRID, as an expert system, is the responsible of monitoring equipment and network applications and taking the adequate decisions to restart a service in the earliest way. This project has been done together Elisabet Berbel and Mònica Ferrer. Thus, it is composed by two documents which are distinguished according to each part of the new InGRID functionalities developed. These improvements in the system have been proposed by the operators of Port d’Informació Científica (PIC), based on the use of the InGRID first release in the production scenario. This document describes specifically the design, the implementation and the testing of the new prototype InGRID second release in the Decisions maker module, where the intelligence of InGRID is contained. Nowadays, the implemented InGRID 2.0 expert system is being used in the production scenario of PIC, providing the operator a reliable and useful tool capable of solving Grid incidences during 24 hours, 7 days a week (24x7).
|Expert System i ... n Grid nodes incidents.pdf||1.648Mb||Restricted access|