Managing Prometheus Instance Metrics

Prerequisites

Your service has been connected for Prometheus monitoring. For details, see Managing Prometheus Instances.

Constraints

• Only the default/common Prometheus instance, and Prometheus instance for CCE/ECS/cloud services/APM support the functions of checking/adding/discarding metrics.
• On the Metric Management page, you can query only the metrics reported in the last three hours.
• Default Prometheus instance: Metrics whose names start with aom_ or apm_ cannot be discarded.
• Prometheus instances for ECS: Only the metrics collected through collection tasks delivered by UniAgent can be displayed.
• Prometheus instances for CCE:

  Only the metrics reported by kube-prometheus-stack (later than 3.9.0) installed on CCE Add-ons or AOM Prometheus instance for CCE Integration Center can be discarded. Ensure that this add-on is running when discarding metrics.

  To view the kube-prometheus-stack status, log in to the CCE console and access the cluster page, choose Add-ons in the navigation pane, and locate that add-on on the right.

Viewing Prometheus Instance Metrics

Only the default/common Prometheus instance, and Prometheus instance for CCE/ECS/APM/cloud services support the functions of checking metrics.

1. Log in to the AOM 2.0 console.
2. In the navigation pane on the left, choose Prometheus Monitoring > Instances.
3. In the instance list, click a desired Prometheus instance. The instance details page is displayed.
4. In the navigation pane on the left, choose Metric Management. On the Metrics tab page, view the names and types of metrics of the current Prometheus instance.
   • Prometheus instance for CCE: You can filter metrics by cluster name, job name, or metric type, or enter a metric name keyword for fuzzy search.
   • Prometheus instance for cloud services: You can filter metrics by metric type, or enter a metric name keyword for fuzzy search.
   • Prometheus instance for ECS: You can filter metrics by metric type, plug-in type, or collection task, or enter a metric name keyword for fuzzy search.
   • Default Prometheus instance: You can filter metrics by metric type, or enter a metric name keyword for fuzzy search.
   • Common Prometheus instance: You can filter metrics by metric type, or enter a metric name keyword for fuzzy search.
   • Prometheus instance for APM: You can filter metrics by environment or metric type, or enter a metric name keyword for fuzzy search.

   Table 1 Metric parameters

   Parameter	Description
   Metric Name	Name of a metric.
   Metric Type	Type of a metric. Options: Basic metric and Custom metric.
   Metrics in Last 10 Min	Number of metrics that are stored in the last 10 minutes. This parameter is not supported for Prometheus instances for cloud services.
   Proportion	Number of a certain type of metrics/Total number of metrics This parameter is not supported for Prometheus instances for cloud services.

Discarding Prometheus Instance Metrics

If Prometheus instance metrics do not need to be reported, discard them.

1. Log in to the AOM 2.0 console.
2. In the navigation pane on the left, choose Prometheus Monitoring > Instances.
3. In the instance list, click a desired Prometheus instance. The instance details page is displayed.
4. In the navigation pane, choose Metric Management.
5. Perform the following operations to discard metrics:
   • To discard a metric, locate it and click in the Operation column.
   • To discard one or more metrics, select them and click Delete in the displayed dialog box.

Adding Prometheus Instance Metrics

If you want to restore metrics after discarding them, add them again.

1. Log in to the AOM 2.0 console.
2. In the navigation pane on the left, choose Prometheus Monitoring > Instances.
3. In the instance list, click a desired Prometheus instance. The instance details page is displayed.
4. In the navigation pane, choose Metric Management.
5. Click Add Metric. In the displayed dialog box, select one or more metrics to restore and click OK.