Generic Cluster
Last updated: 2025-07-15 17:42:40
This document primarily introduces the billing methods, components, and other related information for the TDMQ RocketMQ Edition generic cluster.
Billing Mode
Item | Billing mode | Billing description |
Cluster instance | Monthly subscription - prepaid | When you purchase a generic cluster, the system will calculate the expense bill based on the cluster specifications and purchase duration you selected. You need to settle the bill before you can start using the annual and monthly resources. This billing method is suitable for scenarios where the peak business traffic is relatively stable across different times and requires long-term usage. |
Public network bandwidth | Hourly Bandwidth - Postpaid | Billing is based on the public network bandwidth duration used. The payment pattern is postpaid, with settlements every hour. This is suitable for scenarios where the peak business traffic is relatively stable at different times and only requires short-term usage. |
Cluster instance pricing
Billable Items
The total cost calculation method for purchasing the TDMQ RocketMQ Edition generic cluster is as follows:
cluster overall price = Compute configuration price + Storage configuration price.
Billable Item | Description |
Computing configuration | The computing service fee is charged based on TPS. You can self-define the TPS range (with a step of 4000TPS). The compute configuration price changes linearly with the number of nodes. |
Storage configuration | Primarily for Storage service fees, you can self-define Storage space, and the Storage configuration price changes linearly based on the size of the Storage. |
Pricing
The specific price is as displayed on the purchase page configuration. This section mainly introduces the specifications of a general cluster.
Computing configuration
The TPS specification includes the total of message production and consumption. Traffic exceeding the cluster TPS specification will be strictly rate-limited.
The calculation rules for TPS and the "API call frequency" of a virtual cluster are consistent:
Message types: TDMQ for RocketMQ has four types of messages: normal messages, scheduled and delayed messages, transaction messages, and sequential messages. Among these, scheduled and delayed messages, transaction messages, and sequential messages are all considered advanced feature messages.
Sending or consuming one ordinary message is calculated as 1 TPS, while sending or consuming one Advanced Feature Message (such as Delayed Messages, Transaction messages, etc.) is counted as 5 TPS. For example, if one Topic sends 2 Transaction messages once and consumes one Transaction message, the TPS is 2×5+1×5=15.
Message Size: The upper limit for a single message size is 4MB, measured in 4 KB units. Messages less than 4KB are calculated as 4KB.
Specification Type | Minimum TPS Specification for Single Cluster (4KB Calculation) | Maximum TPS Specification for Single Cluster (4KB Calculation) | Selection Step Size (TPS) |
Universal cluster | 8000 | 80000 | 4000 |
Note:
If the number of nodes in the current specification does not meet your business volume requirements, you can submit a ticket for support.
Storage configuration
Storage costs = Storage space × Unit storage price.
Universal Cluster provides starting Storage of 200GB. You can select more Storage space based on your business needs.
Billable Item | Price (region: Beijing, Guangzhou, Nanjing, Shanghai, Qingyuan, Chongqing) | Price (region: Hong Kong (China), Virginia, Singapore, Silicon Valley) | Price (region: Shanghai Finance, Shanghai Autonomous Driving Cloud) |
Storage Costs (USD/GB/Month) | 0.1381 | 0.1796 | 0.2210 |
Public Network Bandwidth Price
Clusters of TDMQ for RocketMQ 4.x on the public network support both billing by hour and billing by traffic.
Billing by Hour
Region | Price (Unit: USD/Mbps/hour) | |
| 1Mbps - 5Mbps | 6 Mbps and above |
Guangzhou / Shanghai / Nanjing / Beijing / Shanghai Finance / Chongqing / Qingyuan / Shanghai Self-driving Cloud / Hong Kong (China) | 0.0055 | 0.0193 |
Singapore / Virginia / Silicon Valley | 0.0048 | 0.0166 |
hanghai Autonomous Driving Cloud | 0.0085 | 0.0296 |
Billing by Traffic
Billing is based on the outbound traffic, which is the traffic from CLB to the public network.
To avoid high fees due to sudden traffic burst, you can set an upper limit for the bandwidth. If the upper limit is exceeded, packets will be dropped by default, and no fees are incurred.
The traffic units are 1,024-based. For example, 1 TB equals to 1,024 GB, and 1 GB equals to 1,024 MB.
Region | Price (Unit: USD/GB) |
Guangzhou/Shanghai/Nanjing/Beijing/Chongqing/Qingyuan | 0.1429 |
Shanghai Finance | 0.2286 |
Shanghai Autonomous Driving | 0.1714 |
Hong Kong (China)/Singapore/Silicon Valley/Virginia | 0.1857 |
Oversized Topic Charging Rules
Considering the stability of the cluster and real-world usage scenarios, the maximum number of Topics for clusters with different TPS specifications varies. Customers can self-service upgrade to add Topics beyond the free quota limit, and any excess will be charged according to a tiered pricing structure.
Monthly subscription
Oversized Topic Quantity Tiers | Price (USD/Unit/Month, region: Beijing, Guangzhou, Nanjing, Shanghai, Qingyuan, Chongqing) | Price (USD/Unit/Month, region: Hong Kong (China), Virginia, Singapore, Silicon Valley) | Price (USD/Unit/Month, region: Shanghai Finance, Shanghai Autonomous Driving Cloud) |
0-100 | 1.6575 | 2.1547 | 2.3519 |
101-200 | 1.3812 | 1.7956 | 2.2099 |
201-500 | 1.1050 | 1.4365 | 1.7680 |
501-1500 | 0.8287 | 1.0773 | 1.3260 |
1501-2000 | 0.5525 | 0.7182 | 0.8840 |
Above 2,000 | 0.2762 | 0.3591 | 0.4420 |
Billing Example:
Customer A purchased A generic cluster with version 4.x and 8,000 TPS. Users need 1,000 topics. The 8,000 TPS generic cluster supports 400 free topics by default, so the excess number of topics is 600. According to the tiered billing at the top, the additional Topic charges are: 1.6575*100+1.3812*100+1.1050*(600-200) =745.87 USD/month.
Was this page helpful?
You can also Contact Sales or Submit a Ticket for help.
Yes
No
Feedback