What is Amazon Managed Prometheus?
Amazon Managed Prometheus (AMP) is a serverless, Prometheus-compatible monitoring service for container metrics. The service provides the same PromQL query language and compatibility with the Prometheus ecosystem without requiring you to manage Prometheus servers.
Core Features
- Prometheus-Compatible: Use existing PromQL queries and dashboards
- Automatic Scaling: Scales with your metrics without capacity planning
- Long-Term Storage: Metrics are automatically stored for 150 days
- Multi-AZ High Availability: Automatic replication for fault tolerance
- Grafana Integration: Seamless connection to Amazon Managed Grafana
Typical Use Cases
EKS Monitoring: Collect metrics from Kubernetes clusters via the AWS Distro for OpenTelemetry (ADOT) Collector. Pod metrics, node status, and custom metrics are centrally stored and visualized with Grafana.
Microservices Observability: Monitor distributed applications with Prometheus exporters. Each service exposes metrics that AMP centrally aggregates for alerting and analysis.
Hybrid Monitoring: Collect metrics from EKS, ECS, EC2, and on-premises Kubernetes in a central AMP instance.
Benefits
- No Prometheus server management
- Unlimited scaling for metrics
- Compatibility with existing Prometheus tooling
- Integration with AWS security (IAM, VPC)
Integration with innFactory
As an AWS Reseller, innFactory supports you with Amazon Managed Prometheus: collector setup for EKS, alert configuration, Grafana dashboard creation, and migration from self-managed Prometheus.
Typical Use Cases
Frequently Asked Questions
What is Amazon Managed Prometheus?
Amazon Managed Prometheus (AMP) is a managed Prometheus-compatible monitoring service. The service handles scaling, high availability, and long-term storage of metrics. You use standard Prometheus tools like PromQL and Prometheus Exporters.
How does AMP differ from CloudWatch?
AMP is Prometheus-compatible and ideal for container workloads with existing Prometheus tooling. CloudWatch is more deeply integrated with AWS and offers more native integrations. Both can be used in parallel.
How are metrics sent to AMP?
Metrics are sent via Prometheus Remote Write protocol. In EKS, you can configure the ADOT Collector or Prometheus Server with Remote Write backend.
How much does Amazon Managed Prometheus cost?
$0.90 per 10 million metric samples ingested. $0.03 per 10 million samples queried. Storage $0.03 per GB-month. The first 40 billion samples are free in the first year.