imixs-cloud
Monitoring
To get better insights of your Imixs-Cloud project you can install a monitoring service. Kubernetes provides several ways to collect and monitor cluster metrics like the memory or cpu usage of cluster nodes or pods. Kubernetes makes these data available through the Metrics API. These metrics can be accessed either directly by the user with the kubectl top command, or by a controller in the cluster, for example Horizontal Pod Autoscaler, to make decisions.
The Imixs-Cloud monitoring provides different setups to install a monitoring solution.
- metrics-server is a scalable, efficient source of container resource metrics for Kubernetes built-in autoscaling pipelines.
- imixs-prometheus is the recommended monitoring dashboard for Imixs-Cloud
- kube-prometheus provides a configuration setup for a complete cluster monitoring stack based on the Prometheus Operator project.
After a metric server was installed you can monitor your Kubernetes cluster from the commandline tool kubectl top:
$ kubectl top nodes
NAME CPU(cores) CPU% MEMORY(bytes) MEMORY%
master-1 297m 14% 1358Mi 36%
worker-1 1424m 35% 13913Mi 89%
worker-2 1258m 31% 11278Mi 72%
worker-3 1133m 28% 9956Mi 63%
You can also get the data of a singel POD:
$ kubectl top pod traefik-797d34bc7d-l7k8j -n kube-system
NAME CPU(cores) MEMORY(bytes)
traefik-797d34bc7d-l7k8j 22m 50Mi
The k9s tool provides the core functionality to display metrics of the cluster and for each namespace. With the :pulse view you can see node dashboard.
See the install guides for the imixs-prometheus for further details.