Application Real-Time Monitoring Service:[Component upgrade] Multi-tenant monitoring upgrade of Managed Service for Prometheus for cloud products

Overview

Managed Service for Prometheus is upgrading its multi-tenant monitoring feature for cloud products.

What is multi-tenant monitoring for cloud products? This feature allows Alibaba Cloud products such as ApsaraMQ for Kafka and Lindorm to integrate Prometheus capabilities, providing each user with out-of-the-box monitoring dashboards, alerts, and PromQL query services at the product level.

Why is this upgrade happening? As Managed Service for Prometheus migrates from Application Real-Time Monitoring Service (ARMS) to Cloud Monitor, the Alibaba Cloud Observability team is performing a unified upgrade of the multi-tenant monitoring feature to deliver a more stable and feature-rich monitoring experience.

What are the core changes? The multi-tenant monitoring feature previously provided by Managed Service for Prometheus in ARMS stopped receiving version updates in September 2025 and will be migrated to Managed Service for Prometheus in Cloud Monitor.

Comparison of capabilities before and after upgrade

Impact on customers

Data continuity: All existing cloud product monitoring data has been fully migrated. Data from multi-tenant instances of Managed Service for Prometheus in ARMS transitions seamlessly to Cloud Monitor, with no impact on historical data queries.

Network access changes: After the upgrade, only VPC access is supported. Public network endpoints will no longer be available. If you currently rely on public network endpoints, complete your network adjustments before the upgrade.

API permission changes: After the upgrade, RAM users require Cloud Monitor API permissions to access monitoring features. Refer to the attachment for the required permission policy.

Dashboard display changes: The Grafana Shared Edition dashboards will be replaced by Cloud Monitor’s self-developed dashboard engine. The core monitoring experience remains consistent, though some visualization details such as chart styles and layouts may differ.

Alert rule migration: Existing alert rules will automatically migrate to the new version on a per-region basis during the upgrade. Existing ARMS alert rules will not be affected.

Affected products

FAQs

Do I need to perform any manual operations for this upgrade?

In most cases, no. The platform completes the upgrade automatically. However, you must take action in advance if you currently access metric data through public network endpoints, or if your RAM user lacks Cloud Monitor permissions.

Public network access is unavailable after the upgrade. What should I do?

After the upgrade, only access through VPC endpoints is supported. To access from outside the VPC, set up a network connection to the VPC by using VPN Gateway or Cloud Enterprise Network (CEN).

Will my Grafana dashboards still work after the upgrade?

The Grafana Shared Edition dashboards will be replaced by Cloud Monitor’s self-developed dashboard engine. Core monitoring views remain consistent. If you have a self-built Grafana instance, you can continue using it by configuring a new data source endpoint. Integration with Managed Service for Grafana is also supported.

How do I handle permission errors for RAM users?

In the RAM console, attach the least-privilege policy from the attachment to the RAM user. This policy grants the required Cloud Monitor API permissions, replacing the original ARMS permissions.

Do I need to reconfigure alert rules?

No. Existing alert rules will automatically migrate with the upgrade. Integration with the original ARMS Alert Center remains unaffected.

Attachment

• Least-privilege RAM policy for Cloud Monitor Managed Service for Prometheus:

  {
    "Version": "1",
    "Statement": [
      {
        "Effect": "Allow",
        "Action": [
          "cms:GetCmsService",
          "cms:OpenCmsService",
          "cms:CreatePrometheusVirtualInstance",
          "cms:ListPrometheusVirtualInstances",
          "cms:GetPrometheusInstance",
          "cms:ListPrometheusDashboards"
        ],
        "Resource": "*"
      }
    ]
  }