System Monitoring

For all system monitoring and alerting activities, Blue Gecko uses a combination of our own open-source database-monitoring package, Monocle, and the popular open-source package Nagios.

Monocle collects and analyzes data from both Oracle and MySQL databases and stores that data in its own repository for alert processing and trending analysis.  Monocle’s collection routines are lightweight and use as much information that already exists in the database as possible, such as Oracle’s Automatic Workload Repository.  All system monitoring collection is performed locally, and metadata is sent back to our central Nagios cluster for processing over an encrypted channel.

Using custom Nagios plug-ins, the collected data is analyzed and processed for alerting our DBAs, our customers’ DBAs, or any combination.  Schedules and thresholds can be tailored to each customer’s needs and requirements using the robust Nagios alerting engine.  Alerts are sent to our NOC via SMS, email, and are visually represented on the Nagios console.  Acknowledgement of an alert will automatically cut a ticket in the customers’ RT queue, and the issue is worked on the ticket for complete visibility.  If Internet connectivity from our Nagios cluster fails, we use directly connected telephone lines to dial-out and deliver the alerts to our NOC.

Monocle and Nagios allow us to create sophisticated monitors for not only system metrics but application and business metrics as well.  As our monitoring library grows, all customers receive the benefit of new and improved system monitoring.  All alert history is maintained in the Monocle and Nagios repositories.