This document describes the billing mode and billing breakdown of TDMQ RabbitMQ Managed Cluster.
Billing Mode
TDMQ RabbitMQ Open-Source Managed Cluster offers Prepaid monthly subscription and Postpaid by hour billing modes.
Prepaid monthly subscription is a billing mode that requires prepayment to use resources. When you purchase a Managed Cluster, the system will calculate the bill based on the selected cluster specification and duration. You need to settle the bill before starting to use subscription resources. This billing mode is suitable for scenarios with relatively stable peak business traffic at different times and long-term usage.
Postpaid by hour is a payment method billed based on the actual usage of purchased resources. It is mainly suitable for test scenarios or scenarios with uncertain peak traffic. You can use resources first and pay later, with fees settled on an hourly basis at the top of each hour.
Billing Items
The total cost calculation method for purchasing TDMQ RabbitMQ open-source Managed Cluster is as follows:
Overall cluster price = Compute configuration price + Storage configuration price = Price of a single node × Node count + Unit storage price × Storage size.
|
Compute Configuration | The service fees are mainly calculated, with multiple node specifications provided based on demand. There is a minimum node count and a maximum node count range limit. The compute configuration price varies linearly with the node count. |
Storage Configuration | The service fees are mainly for storage. You can customize the storage space (step length is 100GB). The storage configuration price varies linearly with the storage size. |
Public Network Bandwidth | 3 Mbps is provided for free. Fees will be charged for the upgrade section after upgrading public network bandwidth. No fee is charged if public network bandwidth is not upgraded. |
Pricing
Specific fees: The specific price is subject to the cost shown in the purchase page configuration calculation. Estimated cost: If necessary, estimate TDMQ RabbitMQ managed version product configuration cost using the price calculator to quickly generate real-time price.
Price calculator usage reference:
Configuration parameters: Select basic configuration parameters (such as billing mode, region, node specification, node count, storage specification).
Generate budget: Manually input purchase quantity, click add to budget list.
View result: Check billing outcome, click export budget list to obtain the complete configuration detail.
Compute Configuration
Compute configuration fee = Node price × Node count
Note:
If the current specifications' node count limit fails to satisfy your business volume requirement, you can submit a ticket to request support. Warning: Single-node cluster lacks high availability for production and cannot guarantee SLA. Purchase with caution.
Multi-AZ minimum node count: 3
Minimum node count for single availability zone: 1
Maximum selectable nodes: 7 (odd numbers only)
Performance Reference
This section introduces the performance difference of different node specifications in Managed Cluster for your reference when making a purchase.
The following performance is a reference benchmark value we set. The actual performance will not be lower than this specification, subject to the data of user stress testing, and will not generate unexpected elastic capacity fee.
Message type: TDMQ for RabbitMQ supports 4 message types: regular message, persistent messages, priority messaging, and delayed message delivery. Sending or consuming 1 message counts as 1 TPS.
Message size: 4 KB is the measurement unit for 1 TPS. Messages under 4KB count as 4 KB. The upper cap of single message size is 4 MB.
|
| Recommended TPS (production + consumption) | Recommended number of queues | Recommended number of connections |
2-core 8 GB | 750 ~ 1250 | 100 | 1000 |
4-core 16 GB | 2400 ~ 4000 | 200 | 2500 |
8-core 32 GB | 4500 ~ 7500 | 500 | 5000 |
16-core 32 GB | 9000 ~ 15000 | 800 | 8000 |
16-core 64 GB | 9600 ~ 16000 | 1000 | 9000 |
2-core 4 GB (will already be disabled) | 600 ~ 1000 | 100 | 500 |
4-core 8 GB (will already be disabled) | 1500 ~ 2500 | 100 | 2000 |
8-core 16 GB (sold out)
| 3000 ~ 5000
| 200
| 3000
|
4-core 12 GB (sold out)
| 2100 ~ 3500
| 200
| 2500
|
8-core 24 GB (sold out)
| 4200 ~ 7000
| 300
| 4000
|
12-core 24 GB (sold out)
| 6000 ~ 10000
| 400
| 5000
|
Note on single node configuration recommendations:
Message Queue RabbitMQ provides TPS configuration suggestions for reference when purchasing node specifications. Business production environments should rely on actual performance testing data.
When the production consumption TPS of the cluster reaches the recommended value limit, it avoids triggering traffic throttling. However, to ensure the cluster is running normally, it is advisable to use the cluster according to the suggested TPS value to prevent CPU and memory load issues caused by heavy traffic.
- Single-node clusters are recommended for testing scenarios. For production business, select multi-node clusters to ensure high availability of services.
Note: About node specifications "sold out" and "will already be disabled":
The out-of-stock node specification means it is unable to select for new purchase or upgrade on the purchase page. However, the existing cluster control feature, renewal, and upgrade are not affected. It is recommended to use the node specification on sale.
- Specifications that will already be disabled will sell out in the second half of 2025. After selling out, you will not be able to make new purchases or change to corresponding node specifications. However, existing Cluster Management, renewal, and upgrades are not affected.
Storage Configuration
Storage fee = storage space × storage unit price.
Managed Cluster versions all start with 200GB single-node storage. Based on your business requirement, you can select higher storage space. Selection step size is 100GB.
Public Network Bandwidth
In the annual and monthly billing mode, the public network bandwidth cost is as follows:
|
| Portion with a bandwidth of greater than 3 Mbps |
|
| 4Mbps | 5Mbps | 6 Mbps and above (n is the configured bandwidth limit) |
Guangzhou, Shenzhen Finance, Shanghai, Shanghai Finance, Nanjing, Beijing, Chengdu, Chongqing | 0 |
|
| 13.52 | 17.61 | 28.87 + 11.27 × (n - 6) |
Shanghai Autonomous Driving Cloud | 0 |
|
| 20.28 | 26.41 | 43.31 + 16.9 × (n - 6) |
Silicon Valley, Hong Kong (China), Virginia, Taipei (China) | 0 |
|
| 16.9 | 21.13 | 35.21 + 14.08 × (n - 6) |
Frankfurt, Singapore | 0 |
|
| 12.96 | 16.2 | 27.46 + 11.27 × (n - 6) |
Bangkok, Seoul, Tokyo | 0 |
|
| 14.08 | 17.61 | 29.44 + 11.83 × (n - 6) |
Jakarta | 0 |
|
| 14.08 | 17.61 | 28.87 + 11.27 × (n - 6) |
In the pay-as-you-go billing mode, the public network bandwidth cost is as follows:
|
| Portion with a bandwidth of greater than 3 Mbps |
|
| 4Mbps | 5Mbps | 6 Mbps and above (n is the configured bandwidth limit) |
Guangzhou, Shenzhen Finance, Shanghai, Shanghai Finance, Nanjing, Beijing, Chengdu, Chongqing | 0 |
|
| 0.0370 | 0.0482 | 0.0834 + 0.0352 × (n - 6) |
Shanghai Autonomous Driving Cloud | 0 |
|
| 0.0555 | 0.0723 | 0.1252 + 0.0528 × (n - 6) |
Silicon Valley, Hong Kong (China), Virginia, Taipei (China) | 0 |
|
| 0.0463 | 0.0579 | 0.1019 + 0.0440 × (n - 6) |
Frankfurt, Singapore | 0 |
|
| 0.0355 | 0.0444 | 0.0796 + 0.0352 × (n - 6) |
Bangkok, Seoul, Tokyo | 0 |
|
| 0.0386 | 0.0482 | 0.0852 + 0.0370 × (n - 6) |
Jakarta | 0 |
|
| 0.0555 | 0.0723 | 0.1252 + 0.0528 × (n - 6) |