Help Center> Graph Engine Service> User Guide> O&M Monitoring and Alarm Reporting> Graph Instance O&M Monitoring
Updated on 2023-12-20 GMT+08:00
View PDF

Graph Instance O&M Monitoring

GES offers a multi-dimensional O&M monitoring interface that guarantees the smooth operations of graph instances. This feature gathers, monitors, and analyzes disk, network, and OS metrics utilized by graph instances, along with key cluster performance metrics. It promptly identifies significant database faults and performance issues and provides recommendations to optimize and resolve them.

  • The graph instance O&M monitoring dashboard supports only graphs of version 2.3.17 or later.
  • The ten-thousand-edge size is for development learning and does not support the O&M monitoring dashboard.
  • Database edition graph databases do not support graph instance monitoring.

O&M Monitoring Page

  1. Log in to the GES management console. In the navigation pane on the left, choose Graph Management.
  2. In the graph list, locate the target graph instance, click More in the Operation column, and select View Metric to access the Instance Overview page. For details about monitoring metrics, see Monitoring Metrics.
    Figure 1 Instance Overview page

Instance Overview

On the Instance Overview page of a graph instance, you can check the graph instance status, real-time resource consumption, alarm statistics, and service workload. The functions of these areas are as follows:

  • Graph Cluster Status

    In this area, you can check the basic information, cluster capacity, and number of requests of the current graph instance.

    1. Cluster Information: includes graph size and CPU architecture.
    2. Cluster Capacity: includes the number of used and total vertices and edges, as well as the usage.
    3. Cluster Node: includes the number of available/total CNs/DNs.
    4. Cluster Request Statistics: includes the number of waiting read requests, running read requests, waiting write requests, and running write requests.
    Figure 2 Graph Cluster Status
  • Alarm Statistics

    In this area, you can check all alarms that are not cleared for the current instance and all alarms generated for the instance in the last seven days.

    Figure 3 Alarm Statistics
  • Instance Resources

    In this area, you can check the resource usage of the current instance, including the CPU usage, disk I/O, disk usage, memory usage, and network I/O. You can click a resource metric to view its change trend in the last 72 hours and the top 5 nodes with the highest usage of the resource at the current time.

    Figure 4 Instance Resources
  • Workload

    In this area, you can check the change trend of the database service load metric QPS in the last 72 hours.

    Figure 5 Workload