containers and provide to Heapster. The performance data can be stored in InfluxDB and visualized by Grafana, or it can be fed to Google Cloud monitoring for storage and visualization. 1.3.10 Log Management KUBERNETES WITH RANCHER • Ranch-eye: is an haproxy and exposes cAdvisor stats to Prometheus • Grafana: visualization for data • InfluxDB: time series database specifically used to store data from available in Grafana to see what data is available. Note that you will have to use the IP for the host where Grafana container is deployed and NodePort for the Grafana service. Grafana has five distinct

services using telemetry and industry standard monitoring and presentation tools such as Prometheus and Grafana • Support for both CRI-O and containerd runtimes • Support for container and VM workloads (using methods. You can also install the optional Contrail Analytics package, which packages Prometheus, Grafana, Fluentd, and other popular open source software together with Contrail telemetry exporters to provide and the CN2 Web UI. Contrail Analytics packages popular open source software such as Prometheus, Grafana, and Fluentd together with CN2 telemetry exporters to provide an industry-standard way for you to

Prometheus and Grafana at the project and cluster levels and installs pre-configured dashboards that immediately enable visibility into cluster operations. Users can access Grafana and see metrics for begin to scrape custom metrics from them. 3.3.4.2 OpenShift OpenShift ships with Prometheus and Grafana activated by default, with pre-configured alerts and dashboards. As of v4.7, cluster admins can of glass view of cluster metrics of OpenShift clusters using popular open source projects like Grafana, Observatorium and Thanos in the background. 3.3.4.3 Tanzu Tanzu doesn’t include monitoring

dashboards Simple metrics displayed only; Grafana and Prometheus required for displaying complex metrics Simple metrics displayed only; Grafana and Prometheus required for displaying complex

应用监控指标采集分析 目前Kubernetes生态比较广泛使用的通用监控方案是Exporters + Prometheus + Grafana，通过各种Exporter采集不同 维度的监控指标，并通过Prometheus支持的数据格式暴露出来， Prometheus定期拉取数据并用Grafana展现，异常情况 使用AlertManager告警。 常用的一些exporter： • node_exporter