Pay-per-Use Billing

Pay-per-use allows you to use resources before paying for them. It is recommended when you do not want to pay in advance or do not need the resources for long. This section describes the billing rules of AOM.

Scenario

Pay-per-use is suitable for applications or services that cannot be interrupted when facing temporary or sudden traffic increases or unpredictable demands, such as e-commerce flash sales, testing, and scientific computing.

Billing Items

You will be billed based on reported custom metrics, metric storage duration, and data dump volume.

**Table 1** Billing items
Billing Item	Description
Reported custom metrics	The number of reported custom metrics refers to the number of sampling points. It is calculated based on the open-source Prometheus data model. A sampling point data record contains the metric name, label set, sampling timestamp, and value. Sampling points are logically organized on a time line. For example, the kube_node_status_allocatable_cpu_cores metric monitors the number of CPU cores that can be allocated to each node in a Kubernetes cluster. If your cluster contains three nodes, the sampling points are distributed on three different timelines. If the sampling period of each node is 15s, the total number of reported sampling point data records within 1 minute can be calculated as follows: 3 x (60/15) = 12. Scenarios where custom metrics are reported: Why Are Additional Fees Incurred After Prometheus Instances Are Connected?
Metric storage duration	By default, metrics can be stored for 15 days for free. You will be billed for excess days of storage.
Data dump volume	Amount of data that is dumped.

Billing Cycle

The system settles the usage-based fees generated on the previous day at a fixed time every day. It generates a bill for the previous day from 13:00 to 22:00 every day, sends the bill to you, and automatically deducts the fees from your account balance. The settlement time varies depending on actual situations. Pay-per-use resources are settled every day (UTC +08:00).

AOM generates a bill for each Prometheus instance. If your cloud account has multiple pay-per-use instances, multiple bills are generated.

Impact of Changing Prometheus Instances

Scenario: Changing Prometheus instance A to B for custom metric reporting

Impact: If you purchase a pay-per-use Prometheus instance to which custom metrics are reported and then change the instance type, the old bill will be ended while a new bill is generated based on the price of the new instance.

If you change the Prometheus instance type multiple times within a day, multiple bills will be generated. Different bills record the billing for different specifications.

For example, you use Prometheus_A at 09:00:00 and change it to Prometheus_B at 09:30:00. There will be two bills generated between 09:00:00 and 10:00:00.

The first bill corresponds to Prometheus_A fees generated during 09:00:00–09:30:00.
The second bill corresponds to Prometheus_B fees generated during 09:30:00–10:00:00.

Arrears Impact

Arrears Alert

The system will deduct fees for pay-per-use resources at the end of each billing cycle. You will be notified by email, SMS, or internal messages when your account is in arrears.

Arrears Impact

If your account is insufficient to pay your amount, your account goes into arrears. However, your resources will not be stopped immediately; instead, they enter the grace period. You will need to pay for the fees incurred during the grace period, which you can see on the Billing & Costs > Billing Center > Overview page of the Huawei Cloud console. Huawei Cloud will automatically deduct this amount when you top up.

If you do not pay the arrears within the grace period, your resources will enter the retention period and become frozen. You cannot perform any operations on the pay-per-use resources during this period.

If you do not pay the arrears after the retention period expires, your resources will be released and deleted, and their data cannot be restored.