DuploCloud platform comes with an option of centralized metrics for Docker containers, Virtual machines as well as various cloud services like ELB, RDS, ECache, ECS, Kafka etc. These metrics are displayed through Grafana which is embedded into the DuploCloud UI. Just like central logging these are not turned on by default but can be setup with a single click.
Monitoring Kubernetes status with the K8s Admin dashboard
Use the K8s Admin dashboard to monitor various statistics and statuses for Kubernetes, including the number and availability of StatefulSets defined for a service.
In the DuploCloud Portal, select Administrator -> Observability -> Metrics.
Click the k8s tab. The K8s Admin dashboard displays.
Under Observability -> Metrics we have the various metrics per Tenant.
While there are 8-10 out-of-box dashboard for various services, one can add their own dashboards and make them appear in Duplo Dashboard through a configuration
Metrics setup comprises of two parts
Control Plane: This comprises of a Grafana service for dashboard and a Prometheus container for fetching VM and container metrics. Cloud service metrics are directly pulled by Grafana from AWS without requiring Prometheus.
To enable Metrics go under Administrator -> Observability -> Settings. Select the Monitoring tab and click on "Enable Monitoring"
Metrics Collector: Once Metrics control plane is ready i.e. Grafana and Prometheus service has been deployed and are active, you can enable Metrics on a per tenant basis. Navigate to Administrator -> Observability -> Settings. Select the Monitoring tab, and using the toggle buttons to enable monitoring for individual Tenants. This triggers the deployment of Node Exporter and CAdvvisor container in each Host in the tenant similar to how Log Collectors like File beat were deployed for fetching central logs and sending to Open Search.
