Monitoring and Logging
CORD is usually configured with monitoring and logging capabilities, built around Prometheus and ELK Stack, respectively. Both consume events/metrics/logs posted to Kafka. The data they collect can be viewed on their respective dashboards:
Metrics: Collected by Prometheus and viewable using Grafana at port 31300.
Logs: Collected by Elk Stack and viewable using Kibana at port 30601.
The following figure give a high-level overview of the monitoring and logging system.
The monitoring and logging pipelines both share a Kafka bus, with
event topics assigned accordingly (e.g., component.kpis.* and
component.log.*). Components also post other events to Kafka,
but their topics are currently ad hoc and rapidly evolving.
Rationalizing the event schema is on the near-term roadmap.