Managed Service for Prometheus uses pay-as-you-go billing based on actual usage, with free monthly quotas. Resource plans and savings plans are not available on the international site. If you require assistance, submit a ticket.
Billing method
Managed Service for Prometheus uses the pay-as-you-go billing method. Resource plans are not available.
Pay-as-you-go billing is based on actual usage of each billable item, with payment following usage. Free quotas are provided each month.
The international site does not offer savings plans. If you require assistance, please submit a ticket.
Billing structure
Managed Service for Prometheus charges two types of fees:
Prometheus instance fees -- charged for the managed Prometheus service.
Container monitoring fees -- charged for Container Monitoring. These fees include cluster scale fees and Prometheus instance fees.
Prometheus instance fees
Prometheus instance fees cover the following billable items:
Container monitoring fees
Managed Service for Prometheus offers ready-to-use Container Monitoring solutions for Alibaba Cloud container services, featuring two editions: Basic Edition and Pro Edition.
Only ACK Pro clusters support Container Monitoring Pro Edition. Other container cluster types use Container Monitoring Basic Edition by default.
Metric types and volume calculation
Managed Service for Prometheus classifies ingested metrics into two types: basic metrics and custom metrics.
Basic metric
A metric whose ingestion is free. Basic metrics come from three sources:
Alibaba Cloud container service monitoring
Managed Service for Prometheus self-monitoring
Realtime Compute for Apache Flink monitoring
For the full list, see Metrics.
Custom metric
Any metric that falls outside the scope of basic metrics. Custom metrics incur ingestion fees, charged by either reporting volume or write volume depending on the billing mode.
Reporting volume
The number of samples (data points) transmitted when metrics are ingested into a Prometheus instance.
Example: The kube_node_status_allocatable_cpu_cores metric tracks allocatable CPU cores per node in a Kubernetes cluster. For a cluster with 3 nodes sampled every 15 seconds, this metric produces 3 time series. The reporting volume per minute is:
3 nodes x (60 s / 15 s) = 12 samples
Write volume
The size of transmitted data before compression when metrics are ingested into a Prometheus instance.
Example: Using the same kube_node_status_allocatable_cpu_cores metric, if each sample is 0.5 KB, the write volume per minute is:
12 samples x 0.5 KB = 6 KB
Billing cycle
Billing follows a daily cycle:
Resource consumption from the previous day is calculated starting at 00:00 the next day.
Bills are typically generated by 08:00 on the deduction day.
The billed amount is deducted automatically if your account balance is sufficient.
Review your bills in the Expenses and Costs console. Check usage in the Managed Service for Prometheus console. For details, see Usage query.