| 541 | The basis of our work is related to GENI resource monitoring, which includes collecting and recording metrics from the infrastructure that makes up the GENI network. Typically data collection and monitoring frameworks are both limited in the scale of their data collection and the complexity of reporting and alerting on collected data. |
| 542 | |
| 543 | Our work aims to create a scalable system capable of collecting metrics from globally distributed resources. In addition, have developed a highly scalable complex event processing (CEP) platform capable of processing general data streams. As data streams through this system it is recorded as time-series data-points in a distributed database capable of storing billions of metrics. Time series data is then used to model new real-time events. |
| 544 | |
| 545 | Simply, if real-world complex events can be both observed and described in a query language, we can react to these events in real-time. The collection of metrics over time can be correlated with undetected events to develop more accurate event models. |
| 546 | |
| 547 | This capabilities of this system would be of interest to those working in large distributed systems such as those participating in large sensor networks, market trading, resource utilization prediction, systems optimization, etc. |
| 548 | |