Docs

Volcengine Managed Service for Prometheus

Notification template

Extension functions and variables

You can use preset alert variables in notification templates. This topic describes the structure and parameters of preset alert variables.

Data structure

This topic demonstrates the variables, parameters, and data structure contained in the alert information of Volcengine Managed Service for Prometheus (VMP) by using the following sample JSON code. This helps you understand common Go template syntax and its usage.

Parameter description

The following table describes the parameters and descriptions of preset variables that can be used in the notification template.

Variable	Type	Sample Value	Description
Region	String	cn-beijing	The region where the alert was triggered. Valid values: cn-beijing: China (Beijing) cn-shanghai: China (Shanghai) cn-guangzhou: China (Guangzhou) tip For more details, refer to Regions and Availability Zones.
AlertingRuleName	String	The CPU utilization is greater than 80%	The name of the alerting rule.
AlertingRuleDescription	String	Alert description	The description of the alerting rule.
AlertingRuleDetailURL	String	https://console.volcanicengine.com/prometheus/***3503d49e/detail	The URL of the alerting rule. This URL can redirect you to the corresponding alerting rule.
Level	String	P0	The alert level. Valid values: P0, P1, and P2.
PromQL	String	rate(cpu_total[1m])	The PromQL statement that is used to trigger the alert in the alerting rule.
Comparator	String	>	The alert comparison condition. Valid values: >: greater than >=: greater than or equal to <: less than <=: less than or equal to ==: equal to !=: not equal to
Threshold	String	0.8	The threshold of the alert.
For	String	3m	The duration of the alert.
Alerts	Object Alerts	-	The details of the alert resource, which is displayed in a nested structure.

Alerts

Variable	Type	Sample Value	Description
Id	String	cdwaihonj434390ad	The alert ID.
Labels	map[string]string	"cluster": "ccmp26s0bjh01jhk860f0", "namespace": "default", "service": "nginx", "pod": "nginx-85996***"""	The details of the failed resource in the alert.
Value	Float64	0.84	The alert value. tip This parameter only applies to the alerting notification templates and does not apply to the alerting restoration notification template.
StartsAt	time.Time	2024-02-02 17:13:32.065 +0800 CST	The start time of the alert.
EndsAt	time.Time	2024-02-02 17:13:32.065 +0800 CST	The end time of the alert. tip This parameter applies to the alerting restoration notification templates and does not apply to the alerting notification template.
DetailURL	String	https://console.volcanicengine.com/prometheus/*/alert/alerting?detail_id=*	The URL of the alert in the console. This URL can redirect you to the corresponding alert.

Extension functions

Function	Description
vkeClusterName	Obtains a cluster name based on the clusterID of a Volcengine Kubernetes Engine (VKE) cluster.	The VKE metrics written through the collection add-on in a VKE cluster will contain the cluster label, whose value is the cluster ID. You can also use the extension function to obtain a cluster name in an alerting notification. You can perform the following configuration to obtain the cluster name: Ensure that the triggered alerts contain the cluster label. For example, you can use sum(up)by(job,cluster) to achieve aggregation by cluster when you configure alerts. In this case, the labels of the alerts will contain the cluster label. When you configure an alerting notification template, use the vkeClusterName function to obtain the cluster name. Line 4 of the following sample code shows how to obtain the cluster name. {{ range $i, $alert := .Alerts -}} {{ if lt $i 5 -}} No. {{ add1 $i }} resource{{ if $alert.Annotations }}: Cluster name: {{ $alert.Labels.cluster \| vkeClusterName }} {{- range $key, $value := $alert.Labels }} {{ $key }}: {{ $value }} {{- end }} Current value: {{ $alert.Value }} Alert ID: {{ $alert.Id }} Alert start time: {{ date "2006-01-02T15:04:05Z07:00" $alert.StartsAt }} --- {{ else if eq $i 5 -}} The details of only top 5 failed resources are displayed. Click [alert content details]({{ $.AlertListConsoleURL }}) to view more details. {{- end }} {{- end }}
arkEndpointName	Obtains an endpoint name based on endpointID.	The metrics written by ModelArk are all added with the ark_endpoint label, whose value is an endpoint ID. You can also use the extension function to obtain an endpoint name in an alerting notification. You can perform the following configuration to obtain the endpoint name: Ensure that the triggered alerts contain the ark_endpoint label. For example, you can use sum by(ark_endpoint,api)(rate(ark_api_proxy_request_total{}[2m])) to achieve aggregation by ark_endpoint when you configure alerts. In this case, the labels of the alerts will contain the ark_endpoint label. When you configure an alerting notification template, use the arkEndpointName function to obtain the endpoint name. Line 4 of the following sample code shows how to obtain the endpoint name. {{ range $i, $alert := .Alerts -}} {{ if lt $i 5 -}} No. {{ add1 $i }} resource{{ if $alert.Annotations }}: Endpoint name: {{ $alert.Labels.ark_endpoint \| arkEndpointName }} {{- range $key, $value := $alert.Labels }} {{ $key }}: {{ $value }} {{- end }} Current value: {{ $alert.Value }} Alert ID: {{ $alert.Id }} Alert start time: {{ date "2006-01-02T15:04:05Z07:00" $alert.StartsAt }} --- {{ else if eq $i 5 -}} The details of only top 5 failed resources are displayed. Click [alert content details]({{ $.AlertListConsoleURL }}) to view more details. {{- end }} {{- end }}

Configuration example

In addition to the preset alerting templates of the system, you can specify custom content of a notification template based on the data structure of alert variables, Go template syntax, and Markdown syntax. The following section shows the configuration example.

warning

The details of failed resources are obtained through the .Alerts.Labels variable, and the value of the Labels parameter may vary based on different resources. In this example, parameters that contain Labels.* such as $alert.Labels.instance) are used for illustration only. In actual configuration, you need to specify the existing Labels variable of resources.

If your notification template includes multiple resource types, you can also choose not to specify a specific Label variable and instead view the details of each failed resource. For the configuration method, refer to the Template example description section of the Creating a notification template topic.

Last updated: 2026.07.10 11:37:26