The messaging computing specification is a required billable item for ApsaraMQ for RocketMQ. It defines the peak transactions per second (TPS) for message sending and receiving on an instance. If your actual TPS usage exceeds the specification limit, the instance is throttled. If you experience occasional small bursts of traffic, enable elastic TPS for the instance. This increases the TPS limit within a defined range. You are charged for the excess TPS beyond the base specification using the pay-as-you-go billing method.
Background information
Elastic TPS is suitable for scenarios where small, occasional traffic bursts occur. You do not need to reserve extra capacity for these bursts. However, elastic TPS is billed at a higher unit price than the base computing specification. If your traffic regularly exceeds the base specification, upgrade your instance type.
If your actual messaging TPS exceeds the purchased specification limit:
If elastic TPS is enabled, the instance runs normally for TPS values between the base and elastic limits. You are charged for this excess TPS on a pay-as-you-go basis. If usage exceeds the elastic limit, the instance is throttled.
If elastic TPS is not supported or not enabled, ApsaraMQ for RocketMQ throttles the instance.
Only certain instance types support elastic TPS, and each has a maximum elastic TPS limit. For base and elastic TPS limits by instance type, see Instance types.
Billing rules
Item | Description |
Billable item | You are charged for TPS that exceeds the base computing specification limit. This value represents the maximum TPS increment within a 1 hour billing period. |
Billing method | Pay-as-you-go. You use resources first, then pay. Fees are automatically deducted from your Alibaba Cloud account balance. |
Billing cycle | Billed hourly. Usage less than 1 hour is rounded up to 1 hour. |
Billing formula | Hourly elastic TPS fee = Highest excess TPS in 1 hour × Unit price of elastic TPS |
Unit prices for elastic TPS
Primary instance edition | Region | Unit price (USD per TPS-hour) |
Professional Edition | China (Shanghai), US (Virginia), US (Silicon Valley), Germany (Frankfurt), UK (London) | 0.00093 |
China (Hong Kong), Singapore, Japan (Tokyo), Mexico, US (Atlanta) | 0.00125 | |
UAE (Dubai) | 0.00186 | |
Malaysia (Kuala Lumpur), Indonesia (Jakarta), Philippines (Manila) | 0.00135 | |
Platinum Edition | China (Shanghai), US (Virginia), US (Silicon Valley), Germany (Frankfurt), UK (London) | 0.00186 |
China (Hong Kong), Singapore, Japan (Tokyo), Mexico, US (Atlanta) | 0.00251 | |
UAE (Dubai) | 0.00371 | |
Malaysia (Kuala Lumpur), Indonesia (Jakarta), Philippines (Manila) | 0.00269 | |
Standard Edition | Standard instances do not support elastic scaling. | N/A |
Billing example
Assume that a Professional Edition instance in the China (Shanghai) region has a base message send/receive TPS limit of 4,000 and an elastic limit of 2,000. The instance’s operational status over one hour is as follows:
Time | 1st minute | 2nd minute | 3rd minute | ... | 59 minutes | 60 points |
Actual TPS usage (per second) | 4,100 | 4,200 | 4,500 | < 4,000 | 3,700 | 3,700 |
Elastic TPS Increment (TPS) | 100 | 200 | 500 | 0 | 0 | 0 |
During the 1-hour billing cycle, the highest elastic TPS value is 500. The elastic TPS fee for this hour is CNY 3.00. The fee is calculated as follows: 500 × 0.00093 = 0.465.
References
For details about billing for messaging computing specifications, see Messaging computing specification fees.
For base and elastic TPS limits by instance type, see Instance types.
To learn about billing methods and full billing components for ApsaraMQ for RocketMQ, see Billing overview.