View database performance

OCP allows you to view performance monitoring data of a cluster from different dimensions. You can view the data through the following two methods:

• Method 1: Log in to the OCP console, and click Performance Monitoring in the left-side navigation pane.

• Method 2: Log in to the OCP console, and click Performance Monitoring in the left-side navigation pane. The system automatically displays the OceanBase Cluster tab.

Prerequisites

• If you view the database performance monitoring data of a cluster by using Method 1, make sure that the user who logs in to OCP has the following permissions:

  • The Cluster Read-Only resource permission.
  • The Performance Monitoring menu permission.

• If you view the database performance monitoring data of a cluster by using Method 2, make sure that the user who logs in to OCP has the following permissions:

  • The Cluster Read-Only, Tenant Read-Only, or Host Read-Only resource permission.
  • The All menu permission.

Limitations

• For the SQL Response Time monitoring metric:

  • The sub-metric items 99-all and 95-all apply to OceanBase Database of a version in [V4.2.1.7, V4.2.2.0).
  • The sub-metric items commit and rollback apply to OceanBase Database of a version in [V4.2.5.3, V4.3.0.0) or [V4.3.5.2, +∞).

• For the QPS monitoring metric, the sub-metric items commit and rollback apply to OceanBase Database of a version in [V4.2.5.3, V4.3.0.0) or [V4.3.5.2, +∞).

Procedure

The following procedure uses Method 2 as an example.

1. Log in to the OCP console.

2. In the left-side navigation pane, click Performance Monitoring.

3. The system automatically displays the OceanBase Cluster tab.

4. In the Data Filtering section, configure the monitoring data filtering conditions.

   The following table describes the filtering conditions.

   Condition	Description
   Refresh	Refreshes the monitoring data. Auto-refresh is supported. Auto-refresh is disabled by default. If auto-refresh is enabled, you can configure the refresh frequency.
   Select Time (also known as the Monitoring Window)	This option is not displayed in auto-refresh mode. Select a time range to obtain data within this range.
   Cluster	Select a cluster to monitor it.
   Zone	By default, all zones are monitored.
   OBServer	By default, all OBServer nodes are monitored.

5. The system automatically displays the Database Performance tab, where you can view the database performance monitoring data of the cluster.

   • OCP supports viewing database performance monitoring data in four dimensions: cluster, tenant, OBServer, and unit4. By default, data is displayed in the cluster dimension. The monitoring metrics vary depending on the dimension.
   • By default, all monitoring items are displayed in the current dimension. If you do not need to monitor a specific item, you can click Selected Chart on the right to uncheck the monitoring item.
   • All monitoring items support displaying the maximum value (max), average value (avg), and current value (current) of each sub-metric.
   • For more information about the specific meanings and value sources of all monitoring items, see Overview of monitoring metrics.

   The following features help you view and analyze performance monitoring data on the trend chart. You can combine the image and the following description for a better understanding.

   1. View and hide data of other sub-metrics: Hover the pointer over the metric section and scroll up and down with the mouse wheel to view data of all sub-metrics on the trend chart. Click the identifier of another sub-metric to gray it out, and the trend chart no longer displays the information of this sub-metric.

   2. Hover the pointer over a specific point on the trend chart to view detailed data at that point. Drag the pointer to select a region, and the system will zoom in to display the trend information in this region. You can view the detailed trend of this metric in this time period. You can double-click the trend chart to restore it.

   3. Zoom in on a trend chart: Click the zoom-in icon in the upper-right corner of the trend chart. The ① and ② features in the preceding figure also apply to the zoomed-in trend chart.

   4. Edit the monitoring metric: Click the edit icon in the upper-right corner of the trend chart to display only the sub-metrics you want to monitor.

   5. View monitoring data of different dimensions: On the Cluster tab, if you find that the monitoring data of a cluster is abnormal, such as high SQL response time, you can use drill-down analysis to locate the server causing the abnormal response time. Click the Drill-down Analysis icon in the upper-right corner of the trend chart. Supported dimensions for drill-down analysis are: Cluster > Tenant > OBServer.

      The ① and ② features in the preceding figure also apply to the trend chart on the drill-down page. In addition, you can:

      1. On the drill-down page, you can reconfigure the filtering conditions. If you do not set any filtering conditions, they will automatically inherit from the upper-level settings.

      2. If this metric has sub-metrics, you can select a sub-metric to view it, or click ① or ② in the figure to drill down to different dimensions for further analysis.

      3. If this metric involves tenant or host performance, the trend chart provides a direct jump to the relevant tenant or host Monitoring page.

         • To jump to the Performance Monitoring page of a tenant, hover the pointer over the Go to Tenant Performance Monitoring link in the upper-right corner of the trend chart, select the desired tenant, and click it.

         • To jump to the Performance Monitoring page of a host, hover the pointer over the Go to Host Performance Monitoring link in the upper-right corner of the trend chart, select the desired host, and click it.

6. (Optional) Configure the monitoring data collection interval.

   By default, OCP collects monitoring data every second. This high-frequency collection places high demands on the performance of the OCP server and the storage performance of MonitorDB. If the monitoring data does not require high-frequency collection, you can modify the value of ocp.metric.collect.interval.second in config_properties to configure the monitoring data collection interval, thereby reducing resource usage and increasing the overall management scale.

   Note

   Currently, only the values 1, 5, 10, and 15 are supported. If you enter a value that is not listed above, the system will automatically adapt it to the nearest supported value:

   • Any value less than 1 is treated as 1.
   • Any value in the [1,5) range is treated as 1.
   • Any value in the [5,10) range is treated as 5.
   • Any value in the [10,15) range is treated as 10.
   • All other values are treated as 15.

   After you configure the monitoring data collection interval, OCP immediately starts collecting data at the new interval. The display interval of the monitoring data is three times the collection interval (except for 1 second). If the collection interval is 5 seconds, the display interval of the monitoring data is 15 seconds.