Enterprise Distributed Application Service (EDAS) integrates with Application Real-Time Monitoring Service (ARMS) so that you can monitor key performance metrics and manage alerts for your applications that are deployed to EDAS.

Implement application monitoring

EDAS integrates with ARMS so that you can monitor key performance metrics for your applications that are deployed to EDAS. This helps you locate interfaces that have errors or slowly respond and make the called parameters reoccur. Therefore, the efficiency of problem diagnostics in the production environment is significantly improved.

Monitoring granularity Description References
Application Overview On the Application Overview page of an application, you can view the information about the related Kubernetes cluster, such as the region, microservice namespace, and status of the cluster, and the application diagnosis report. You can also view key metrics for the health status of the application. The metrics are classified into overall metrics, metrics of provided Services and dependent Services, and system metrics. Overall metrics include the total number of requests and average response time. System metrics include CPU utilization and memory usage. View the overall information about an application
Prometheus The preset monitoring dashboards that are provided by the Prometheus monitoring feature show the basic information, CPU metrics, memory metrics, and network metrics of pods. On these dashboards, you can view the Prometheus monitoring metrics, and change the properties of dashboard data as needed, such as time intervals and refresh rate. View Prometheus monitoring metrics
Instance Details EDAS provides various application monitoring metrics:
  • The Java Virtual Machine (JVM) monitoring feature is used to monitor important JVM metrics, such as heap memory, non-heap memory, direct buffer, memory-mapped buffer, garbage collections (GCs), and JVM threads.
  • JVM monitoring can intuitively display multiple memory metrics within a specified period of time. However, although the charts can reflect excessive memory usage, specific information cannot be displayed. Therefore, the charts cannot help you troubleshoot problems. To resolve this issue, you can create a memory snapshot task and obtain detailed information about memory usage from logs.
  • The host monitoring feature is used to monitor the metrics of the CPU, memory, disks, load, network traffic, and network packets.
Service Details This feature is used to monitor the details of interface calls of an application, including SQL analysis, NoSQL analysis, exception analysis, error analysis, upstream and downstream services, and interface snapshots. Service and API monitoring
Application Diagnosis - Real-time Diagnosis This feature is applicable to scenarios in which you need to monitor application performance and identify the causes of problems in a short period of time. Real-time diagnostics
Application Diagnosis - Exceptions Diagnosis The application exception analysis feature collects statistics on the number of exceptions, the number of exceptions of each type, and the ports on which exceptions occur. Exception analysis
Application Diagnosis - Threads Profiling This feature provides statistics on the CPU time consumption at the thread level and the number of threads of each type. The method stacks of threads are recorded and aggregated every 5 minutes. This helps you review the code execution process and locate thread issues. Thread profiling

Alerts

You can customize alert rules for specific monitored objects. If a rule is triggered, the system sends alert messages to the specified contact group by using the notification method that you specify. This way, the contacts can resolve issues at the earliest opportunity.