Overview

Functions

• Multi-dimensional data insights: comprehensively monitor Kubernetes native containers and displays metrics of clusters, nodes, workloads, pods, and events for the health status and load of clusters.
• Dashboard: shows various graphs such as line graphs and digit graphs on the same screen for comprehensive data display.

Advantages

• Monitoring Center is deeply integrated with Prometheus, a mature monitoring project of the Cloud Native Computing Foundation (CNCF). It brings in observability for your cloud native applications by collecting, storing, and visually presenting O&M data, such as key metrics and events.
• Monitoring Center provides full-stack monitoring from cloud native infrastructure resources to application workloads, enabling you to clearly perceive the infrastructure and application load status anytime anywhere.
• Monitoring Center monitors Kubernetes clusters, nodes, and pods, enables end-to-end tracing and visualization for services, and provides cluster health diagnosis, greatly speeding up fault analysis and locating.
• Monitoring Center provides ready-to-use add-ons, data collection, and cloud native cluster monitoring. Compared with monitoring developed based on open source components, it is more competitive in reliability, availability, and deployment.
• Monitoring Center provides lightweight add-ons for metric collection. Compared with Prometheus, it greatly reduces resource usage and facilitates deployment.

Monitoring Center Architecture

Figure 1 Monitoring Center architecture

The cloud native cluster monitoring add-on collects metrics exposed by exporters in CCE clusters and writes the data to the AOM instance in Prometheus RemoteWrite mode.

Monitoring Center provides multi-dimensional data insights and dashboard based on the metrics stored in the AOM instance.

Based on the RemoteWrite feature, the cloud native cluster monitoring add-on reports monitoring metrics in a cluster to the third-party monitoring platform through Bearer Token authentication.

Prometheus Monitoring

Prometheus has become the most common tool for cloud native observability. Its powerful monitoring capability and active community ecosystem enable Prometheus projects to be hosted under the CNCF. Currently, the CCE add-on page provides an add-on (Cloud Native Cluster Monitoring) for monitoring Kubernetes clusters.

The cloud-native monitoring add-on supports Prometheus Server and Prometheus Agent (a lightweight version of Prometheus Server). Prometheus Agent provides a lightweight metric collector for collecting monitoring metrics. Prometheus Server provides thanos-based HA to ensure that Prometheus can run normally in the case of a single point of failure.

Based on the Prometheus monitoring ecosystem, AOM provides hosted Prometheus instances for CCE, which are suitable for monitoring CCE clusters and applications running on them. By default, AOM instances integrate the cloud-native monitoring add-ons of CCE clusters. After Monitoring Center is enabled, metrics are automatically reported to the specified AOM instances.

AOM ICAgent Monitoring

As the collector of AOM, ICAgent runs on hosts to collect metrics, logs, and application performance data in real time. For hosts purchased from the ECS or BMS console, manually install the ICAgent. For cluster nodes, ICAgent is automatically installed.