Skip to main content

Monitoring

This article describes how to view service-related monitoring data.

Operation Steps​

  1. Enter Cloud Hosting and select the service you want to query from the service list.
  2. On the service page, you can view the basic monitoring overview. By clicking More Details, you can access the monitoring page of operations management. Alternatively, by directly selecting the operations management of the environment, you can switch to the monitoring interface to view cloud hosting monitoring information.
  3. In the Operations Dashboard of Operations Management, select Cloud Hosting from the Monitoring Items, then expand to select the specific service you want to query.
  4. In the dropdown list, you can select monitoring for specific service dimensions, or further select the version of a specific service to view version monitoring.

Monitoring Field Descriptions​

  • In the Service/Version Dimension Monitoring tab, the top time filter applies to all monitoring data on the page. For example: selecting "7 days" at the top will display data from the past 7 days in both statistical cards and curves.
  • In monitoring graphs, "granularity" refers to the unit of time corresponding to each monitoring data point. It varies with the selected time range: longer time spans correspond to coarser granularity. All values on the graph must be interpreted in conjunction with the granularity.

Service Monitoring​

  • Invocation count: The total number of requests received by the service per unit time (including requests generated through service domains and HTTP access services).
  • Response Time (ms): The average response time of all requests received by the service per unit time.
  • QPS: The average number of requests processed per second by the service within a unit time.
  • Error Responses: The number of HTTP errors (such as 404, 500) returned due to failed service requests per unit time.
  • CPU Usage: The sum of CPU resource consumption across all versions of the service per unit time, measured in (core x hours).
  • Memory Usage: The sum of memory resource consumption across all versions of the service per unit time, measured in (GiB x hours).
  • Instance Count: The average value per unit time of the sum of real-time instances across all versions of the service.
  • Abnormal Instance Count: The average value per unit time of the sum of real-time instances in abnormal status across all versions of the service. All instances not running normally are marked as abnormal, including those in the startup phase.

Version Monitoring​

  • Invocation count: The total number of requests received by the version per unit time (including requests generated through service domains and HTTP access services).
  • Response Time (ms): The average response time of all requests received by the version per unit time.
  • QPS: The average number of requests processed per second by the version within a unit time.
  • HTTP Errors: The number of HTTP errors returned due to failed requests to the version per unit time.
  • CPU Usage: CPU resource consumption of the version per unit time, measured in (core x hours).
  • Memory Usage: The sum of memory resource consumption of the version per unit time, measured in (GiB x hours).
  • CPU Utilization: The average CPU utilization across all instances of the version, further averaged per unit time. This can serve as a reference value for setting scaling conditions.
  • Memory Utilization: The average memory utilization across all instances of the version, further averaged per unit time.
  • Instance Count: The real-time instance count of the version, averaged per unit time.
  • Abnormal Instance Count: The average number of real-time instances in abnormal status per unit time. All instances not running normally are marked as abnormal, including those in the startup phase.