Troubleshooting - monitoring perspective

Gaining confidence in resolving issues when operating a large infrastructure based on OpenStack should be achieved in the first stage by an efficient monitoring and alerting system. The second round of troubleshooting tasks can be performed by digging into log files and debug messages thrown by the system or some parts of it. We have already set up a monitoring system that acts as a first watcher on our overall OpenStack installation. Reacting to escalated alerts generated by monitoring systems will first point to specific components of OpenStack that would help to start gathering valuable information for further investigation. Most importantly, the monitoring tool must be capable of triggering human readable notifications when operators and administrators configure a certain number of events that are highly customized. This will help quickly identify not only the affected...