Enabling Time Series Data Stream

When Kafka is used as part of an infrastructure, multiple vovservers can now be enabled to provide time series data and events, without negatively impacting server performance/scalability. This allows users to capture time varying data in order to see how usage evolves over time.

Events Frequency

The metrics events are published at the same rate that metrics are calculated in the server. This will vary by load but should be at most every 10s on an active server, and possibly longer if the server is heavily loaded such that the scheduling cycle takes longer than this. The project data is expected to be relatively static and is published every 4 hours. The command: vovservermgr config sds.readconfig 1 will cause the updated project record to be published on execution.

Project IDs

To enable data from multiple projects to be collected on the same kafka infrastructure, each event will contain the field projId which identifies the project which published the event. The projId field is formed by concatenating the project name, a hyphen, and the vovserver instance's numeric generated unique id. e.g. "vnc-12345678"