A failure prediction system could be deployed to help the NFV system avoid the unexpected failure in advance. The whole failure prediction system is made up of a data collector, a failure predictor and a failure management module.
This project aims to development a predictor framework, a data collector model and a dashboard. User could define which failure should be detected and which data should be used for prediction via dashboard.
The predictor framework is based on machine learning library, e.g. Weka, Spark ML lib. User can define what failure should be detected and which data is used for prediction. The predictor will collect data automatically and perform corresponding prediction.
The data collector consists of Ceilometer and Monasca which can be extended to plugin some other open source data collectors, e.g. Zabbix, Nagios, Cacti. Based on real-time analytics techniques and machine learning techniques, the failure predictor analyses the data gathered by the data collector to automatically determine whether a failure will happen. If a failure is judged, then the failure predictor sends failure notifications to the failure management module, which could handle these notifications.
In Brahmaputra release, we have defined requirement documents. In release Colorado, we will deliver code, including dashboard, data collector and predictor framework.
Prediction's data collection model can collect data through Ceilometer, Zabbix and Monasca. User could select metrics in scope of Ceilometer, Zabbix and Monasca for prediction. All metrics are defined below.
[prediction]in the subject for easier filtering