A system architecture based on ACK has four observability layers. From the bottom up, these layers are infrastructure, container performance, application performance, and user business. The monitoring architecture of ACK Edge is the same as that of ACK, but their monitoring capabilities differ. This topic describes the observability capabilities of each layer in an ACK Edge cluster.
Infrastructure layer observability
Observability for the infrastructure layer involves monitoring the underlying resources of an ACK Edge cluster. Common scenarios include tracing service invocation links within the resource pools of pods and nodes, visualizing the system topology, and monitoring infrastructure performance. This includes monitoring host nodes and basic network components.
Solution | Description | Scenarios | References |
Architecture visualization solution | In a Kubernetes cluster, services run on a resource pool of multiple nodes. This makes it complex to trace invocation links and topology relationships between pods. Effective monitoring of cluster load status and traffic throughput is crucial for cluster management. The Alibaba Cloud monitoring solution for Kubernetes combines eBPF technology with Prometheus monitoring for containers. It provides an integrated observability solution for Kubernetes clusters. The solution covers metric monitoring, application tracing, log analysis, and event monitoring. It enables network monitoring and architecture visualization for an ACK Edge cluster without changing application code. This gives developers and O&M engineers comprehensive observability to understand and manage cluster resources. | Applicable to all scenarios.
| For more information, see Cluster topology monitoring. |
Container performance layer observability
Observability for the container abstraction layer in systems built on ACK Edge includes monitoring the performance and events of clusters and containers.
Monitoring performance metrics for clusters and containers
Solution | Description | Scenarios | References |
Managed Service for Prometheus solution | A fully managed monitoring service compatible with the open source Prometheus ecosystem. Provides ready-to-use dashboards for a wide range of components without self-managed monitoring infrastructure. Recommended. | Applicable to all scenarios. This includes advanced observability capabilities, such as monitoring metrics of cluster components and custom monitoring. | For more information, see Monitor an ACK Edge cluster with Managed Service for Prometheus. |
Event monitoring for clusters and containers
Solution | Description | Scenarios | References |
Event monitoring solution | Event monitoring is another monitoring method that focuses on Kubernetes events. It can compensate for the shortcomings of resource monitoring in real-time performance, accuracy, and scenario coverage. Developers can retrieve events to diagnose cluster abnormalities and issues in real time. We recommend using the Event Center monitoring feature provided by Alibaba Cloud Simple Log Service (SLS). | Applicable to all scenarios. | For more information, see Event monitoring. |
Application performance layer observability
Observability at the application performance layer covers specific application scenarios in systems built on ACK Edge. This includes monitoring application performance metrics, system call tracing, and logs. For example, for a Java application built on ACK, you can monitor metrics such as the number of threads.
Solution | Description | Scenarios | References |
Non-intrusive Java application monitoring (APM) solution | Application Real-Time Monitoring Service (ARMS) provides intrusion-free APM for Java applications in ACK. Install the ARMS add-on to locate faulty and slow interfaces, tune parameters, detect memory leaks, and identify performance bottlenecks. Recommended. | Java application monitoring. Intrusion-free. | For more information, see Java application monitoring. |
User business layer observability
ACK Edge supports various business scenarios, such as building a high availability (HA) and scalable website with a container service, tracking website operational data such as page views (PV) and unique visitors (UV), and performing application cost audits.
Solution | Description | Scenarios | References |
Custom log monitoring solution | Customize application log content and format, collect logs with SLS, and configure dashboards for business monitoring or system auditing. Recommended. | All scenarios, such as traffic monitoring, cost auditing, and order trend analysis. | For more information, see Collect container logs from an ACK Edge cluster using Simple Log Service. |
Custom business dashboards with Managed Service for Grafana | Managed Service for Grafana is a cloud-native visualization platform with O&M-free runtime environments. Ingests data from Alibaba Cloud services such as databases, Message Queue, Prometheus, and SLS, and provides fine-grained monitoring dashboards. Analyze metrics, logs, and traces without managing servers or updates, with built-in security and high availability. | Applicable to all scenarios. Configure dashboards for business needs, such as real-time PV and UV monitoring. | For more information, see What is Managed Service for Grafana?. |
Use ARMS Browser Monitoring to monitor service traffic and health from the frontend. | Monitors web applications, Weex, and mini-programs by tracking page load speeds, JS error stability, and API call success rates. | Applicable to JavaScript frontend application scenarios. | For more information, see |
References
For more information about Log Monitoring, see Collect container logs from an ACK Edge cluster using Simple Log Service and Collect logs of control plane components in an ACK Edge cluster.
For more information about monitoring metrics, see Java application monitoring, Cluster topology monitoring, and Event monitoring.
For more information about how to use monitoring features and configure dashboards, see Monitor an ACK Edge cluster with Managed Service for Prometheus and Use Prometheus Query Language to query Prometheus monitoring data.