All Products
Document Center


Last Updated:Dec 06, 2023

Managed Service for Prometheus allows you to collect, store, compute, and display monitoring data. You can also set alerts for a wide range of metrics. Managed Service for Prometheus supports monitoring data that is collected from containers, cloud services, middleware, databases, applications, and business. This topic describes the features that are supported by Managed Service for Prometheus.

Integration with monitoring objects



Create a Prometheus instance

Multiple types of Prometheus instance are supported. You can create a Prometheus instance of any type based on your needs.

Health checks

Health check for cloud services, health check for ACK services, and custom health check are supported.

You can use the health check feature to test the connectivity of monitored services on a regular basis. The feature provides the health statuses of the monitored services so that you can detect and handle exceptions at the earliest opportunity.

Metric collection



Service discovery

Default service discovery: This feature is built into Managed Service for Prometheus and is automatically enabled when you use Managed Service for Prometheus.

In default service discovery, the metrics of pods in all namespaces of the Kubernetes cluster are collected.

ServiceMonitor: You can add ServiceMonitors and configure rules to collect metrics for Managed Service for Prometheus.

PodMonitor: You can add PodMonitors and configure rules to collect metrics for Managed Service for Prometheus.

Edit the Prometheus.yaml file

You can edit the Prometheus.yaml file to configure a data collection rule. Managed Service for Prometheus uses the rule to monitor an application.

View metrics

You can view basic metrics and custom metrics.

You can discard metrics that are no longer needed.


On the Targets tab, you can view the targets that are being captured and check the status of each target. You can also view the metrics that are exposed in each target.

Monitoring data processing



View the remote write endpoints

You can use the remote write feature to store the monitoring data of Managed Service for Prometheus to remote databases.

You can use remote read and write endpoints to store the monitoring data collected by self-managed Prometheus systems in Prometheus instance.

View the remote read endpoints

Modify the RecordingRule.yaml file

Recording rules can be used for the custom development of metric data. You can configure recording rules to advance the computing process to the write side to reduce the resource usage on the query side. Especially in large-scale clusters and complex business scenarios, PromQL complexity can be effectively reduced. This improves the query performance and simplifies user configurations.

Global aggregation instance

Provides a virtual aggregation instance for all Prometheus instance in the current region. This virtual aggregation instance can implement unified metric query and alerting.

Display of monitoring data



View Grafana dashboards

A variety of Grafana dashboards are built into Managed Service for Prometheus. You can also customize dashboards to display monitoring data.

View the HTTP API endpoints

Managed Service for Prometheus provides HTTP API endpoints. You can use an HTTP API endpoint to connect a Prometheus instance to your custom Grafana dashboard and obtain the monitoring data of the Managed Service for Prometheus for custom development.




Create an alert rule

A variety of alert rules are built into Managed Service for Prometheus. You can customize alert rules for specific monitoring objects.

If an alert rule is triggered, the system sends notifications to the specified contact group by using the specified notification methods. This way, the alert contacts can resolve issues at the earliest opportunity.

Intelligent detector

Managed Service for Prometheus uses an intelligent detector to identify abnormal fluctuations of KPI time series data. This helps you make informative decisions in automatic data loss protection and root cause analysis if alerts occur.

Prometheus instance management



Change the storage duration for a metric.

You can change the storage duration of the data samples for a metric.

Set the number of agent replicas

Managed Service for Prometheus provides Horizontal Pod Autoscaling (HPA) capabilities that automatically increase or decrease the number of agent replicas. Workloads are distributed across collection tasks. Compared with open source Prometheus, Alibaba Cloud Managed Service for Prometheus provides higher scalability and availability.

Manage a Prometheus agent

You can view the basic information and health check results of the Prometheus agent, set the number of agent replicas, and restart the agent.


The features supported by different types of Prometheus instance may vary.