|
|
|
|
Each component of Traverse operates independently to provide a high level of scalability and fault tolerance. When you start a DGE, it searches the dge.xml file for its unique name, which must match the name you specified when you created the DGE. The DGE connects to the BVE ObjectStore—specified in the emerald.xml file—and downloads the entire configuration associated with that unique name, including tests, thresholds and actions.
After this process completes, the DGE performs tests, generates events when thresholds are crossed, and triggers the corresponding notifications. The data collected by each DGE is stored in a local SQL database on the DGE itself.
Any configuration changes made in the BVE ObjectStore through the BVE API or the Web Application is instantly transmitted to the appropriate DGE.
When a user logs into the Web Application, the system searches the Provisioning Database for the list of devices that the user has permission to view. The Web Application then connects directly to the distributed DGEs and gets the real-time status of the services or devices. When the user needs a report, the Web Application fetches the data using parallel queries from the distributed DGEs and generates the reports in real time.
Log files and SNMP traps are managed by the Input Stream Monitor (ISM) and the trap daemon, respectively, and then matched against user-defined regular expressions that define severities and corresponding actions/notifications. You can view these text-based messages in the Message Window in the BVE Web Application.