How to Seamlessly Migrate Your Self‑Hosted Prometheus + Thanos to Alibaba Cloud Managed Prometheus

Introduction

Prometheus has become the de‑facto standard for observability metrics. Some users still operate self‑built Prometheus + Thanos + Grafana stacks for infrastructure and application monitoring. Alibaba Cloud Managed Prometheus fully integrates the open‑source ecosystem, offering high‑performance, high‑availability, scalable, low‑cost, and easy‑maintenance metric observability. This article discusses migration solutions for typical scenarios.

Typical Self‑Built Deployment Scenarios

Kubernetes cluster installations: Alibaba Cloud Container Service, other cloud providers' container services, self‑built IDC Kubernetes clusters.

Non‑Kubernetes installations: Alibaba Cloud ECS, other cloud providers' ECS, self‑built IDC VMs.

Users often deploy one or more Thanos instances on top of Prometheus for long‑term storage and aggregation.

Pain Points of Self‑Built Prometheus + Thanos

Heavy manual work to install exporters, configure dashboards and alert rules; open‑source Grafana dashboards are often insufficiently optimized.

Multiple installations across different clusters or VPCs increase deployment cost and operational complexity.

Large‑scale clusters overwhelm self‑built Prometheus’s metric ingestion capacity.

Adding Thanos for long‑term storage introduces additional complexity and cost, especially for non‑Kubernetes environments requiring Thanos Receiver.

Alibaba Cloud Managed Prometheus Capabilities

The service fully aligns with the open‑source Prometheus ecosystem, supports a rich set of component observations, provides out‑of‑the‑box dashboards, and offers a fully managed hybrid‑cloud/multi‑cloud Prometheus service. It supports Alibaba Cloud Container Service, self‑built Kubernetes, Remote Write, and hybrid‑cloud/multi‑cloud ECS metrics, as well as multi‑instance aggregation for unified queries, Grafana data sources, and alerts.

Migration Scenarios and Steps

1. Metric Collection Migration

Metric collection means Prometheus pulls and stores metrics from configured jobs and exporters.

Alibaba Cloud Container Service (Container Cluster Monitoring)

Configure and validate Alibaba Cloud Prometheus; wait for the same retention period as the self‑built instance before switching.

In the console’s Access Center, select “Container Cluster Monitoring” to create a container environment instance.

In Access Management, configure each monitoring component to generate collection jobs, default dashboards, and alert rules.

Optional: configure custom ServiceMonitor, PodMonitor, or Job rules.

Verify that the default container‑cluster dashboard and component dashboards are functional.

Self‑Built Kubernetes (Registered Cluster)

Register the self‑built Kubernetes cluster as an Alibaba Cloud Container Service “registered cluster” and follow the same steps as above.

Alibaba Cloud ECS (Host Monitoring)

In Access Center, choose “Host Monitoring” to create an ECS environment and deploy the required exporters.

In Access Management, configure VPC components for collection jobs, dashboards, and alerts.

Optional: add custom collection rules.

Validate the default dashboards for each component.

Self‑Built IDC (VM Monitoring)

Connect the IDC to an Alibaba Cloud VPC via a dedicated line.

Install Node/Process exporters on each VM.

In Access Center, select “Host Monitoring (Self‑install + IP selection)” to create an ECS environment and generate collection configs.

Configure VPC component access in Access Management, add optional custom rules, and verify dashboards.

General RemoteWrite Scenario (Generic Instance)

Create a generic Prometheus instance in the console.

Modify the self‑built Prometheus configuration to Remote Write metrics to the Alibaba Cloud instance.

In the Metrics Center, query and verify the ingested metrics.

2. Visualization Migration

Alibaba Cloud Prometheus provides built‑in professional dashboards for common components (e.g., MySQL, Redis) in container/Kubernetes and ECS scenarios.

Default dashboards are available in Access Management → Dashboard Query.

For custom dashboards, create an Expert‑edition Grafana workspace, then migrate or import the desired dashboards.

In RemoteWrite generic instances, simply switch the Grafana data source to the corresponding Alibaba Cloud Prometheus instance.

For multi‑environment aggregation, use the Global Aggregation Instance capability.

3. Alert Configuration Migration

Professional, out‑of‑the‑box alert rules are also provided for the same components.

View default alert rules in Access Management → Integrated Components → Alert List.

Custom alert rules can be created or imported via the Prometheus Alert Rules or Alert Rule Templates menus.

For aggregated alerting across multiple environments, use the Global Aggregation Instance feature.

About Alibaba Cloud Managed Prometheus

Built on the cloud‑native observability standard Prometheus, the fully managed service integrates common cloud services, supports mainstream open‑source components, and covers business, application, middleware, and system observability. It offers ready‑to‑use Grafana dashboards, intelligent alerts, and unified query via PromQL and SQL, helping enterprises quickly establish a one‑stop metric observability system, reduce operational overhead, and improve reliability.