edit-icon download-icon

Application monitoring overview

Last Updated: Feb 01, 2019

Application monitoring can accurately reflect the real-time traffic and history information of an application, allowing you to monitor the application health and quickly pinpoint issues.

  • You can use EagleEye provided by EDAS for the infrastructure monitoring and service monitoring of applications, and view the dashboard.
  • If you have enabled the advanced monitoring service, you can use ARMS to monitor applications.
  • Meanwhile, for applications deployed in the native Spring Cloud or Dubbo framework later than x x, xxxx and applications deployed in Container Service Kubernetes clusters, ARMS is used for monitoring by default.

Glossary

  • TraceId: corresponds to a request. It is globally unique and transmitted between systems.

    • IP Address: indicates the IP address (hexadecimal) of the ECS instance that creates the TraceId.

    • Creation Time: indicates the link creation time.

    • Order: used for link sampling.

    • Flag Bit (optional): used for debugging and identification.

    • Process ID (optional): used for single-instance multi-process applications.

  • RpcId: calls and identifies the log track order and nesting relationships. This item is transmitted between systems.

  • Service Dimension: Service data is monitored in the application and service dimensions. Data in the application dimension is aggregated by application, while data in the service dimension is aggregated by custom service. For example, application A provides services a, b, and c.

  • Drill Down: views the metrics of upstream (downstream) applications associated with the target metric.

Types of monitoring data

The Service Monitoring page provides the tabs for different data types, enabling pertinent monitoring.

  • RPC Call Overview: displays the RPC services (including the HSF and other customer services) that are provided by an application as the server.
  • RPC Call Source: displays the records of the following applications that call the RPC services provided by the current application.
  • RPC Call Dependency: displays the records of the current application that calls the RPC services (including HSF and other custom services) provided by other applications.

Types of monitoring reports

  • Mix of Graph and Table (Default): displays data in the “table + graph” combination, including the monitored target, time, QPS, response time, server response time, errors, and results. By default, the graph shows the data for the last hour, and the table lists the data for the last five minutes.
  • Multi-graph: displays data in a graph, including the monitored target, time, QPS, response time, errors, and results. By default, this graph shows the data for the last hour and separately lists the latest data.
  • Table: displays data in a table, including the monitored target, QPS, response time, errors, and results. The table lists the data for the last minute.

Metric descriptions

  • Errors per Second: records the RPC error rate per minute, which is the total number of errors that occurred within the minute divided by 60.
  • Results per Second: records the returned result in “Result: QPS” format, where the “Result” indicates the RPC result. The HTTP result is consistent with the HTTP ErrorCode.
Thank you! We've received your feedback.