tencent cloud

Pro Clusters
Last updated: 2025-12-24 14:52:52
Pro Clusters
Last updated: 2025-12-24 14:52:52
This document describes the price of each billable item for TDMQ for Apache Pulsar pro clusters.

Billing Specifications Description

Limits of Cluster Specifications

TPS upper limit for message sending and receiving: the TPS upper limit for message sending and receiving (production and consumption) in TDMQ for Apache Pulsar pro clusters.
Note:
For the specific TPS upper limit corresponding to different cluster specifications, see Pro Cluster Specifications.
Upper limit of peak bandwidth: Taking 45 MB/s as an example, this means the inbound and outbound bandwidth of the TDMQ for Apache Pulsar pro cluster can each reach 45 MB/s, including traffic for 3 replicas.
Other resource limits: Cluster specifications differ not only in TPS and bandwidth capabilities, but also in some management resources and configuration limits. For specific limits of each specification, see Pro Cluster Specifications.

After the Limits of Cluster Specifications Are Exceeded

Pro Cluster - Fixed Storage

If your actual usage exceeds the specification limits of the purchased cluster, the platform will not enforce strict traffic throttling measures. It is recommended that you subscribe to alarms for core cluster metrics, monitor corresponding message processing performance during metric fluctuations, and upgrade the cluster specifications in a timely manner to ensure stable business operations.

Pro Cluster - Elastic Storage

If your actual usage exceeds the specification limits of the purchased cluster, the platform will perform traffic throttling based on the sending-receiving ratio configured during cluster creation. This will affect normal message sending and receiving. It is recommended that you subscribe to alarms for core cluster metrics and upgrade the cluster specifications in a timely manner to ensure stable business operations.

Cluster TPS Calculation Specifications

1. The TPS for message sending and receiving refers to the total number of messages sent and subscribed per second.
2. For messages with advanced features, they can be divided into two cases:
For scheduled/delayed messages, transactional messages, and Key-Share subscriptions, the number of API calls for sending messages needs to be multiplied by a multiplier of 5 compared to general messages. For example, sending 1,000 delayed messages per second results in a message sending TPS of 1,000 x 5 = 5,000 times/second.
For tag messages, production is calculated as normal messages, and during consumption, the metering includes the volume of filtered messages. For example, sending 1,000 messages per second, with a message subscription volume of 2,000 and a message filtering volume of 5,000, results in a message sending and receiving TPS of 1,000 + 2,000 + 5,000 = 8,000 times/second.
3. Message size is measured in 4 KB units, and TPS is calculated with different multipliers based on the message size:
Message Size
Multiplier
N ≤ 4KB
1
4KB < N ≤ 16KB
2
16KB < N ≤ 100KB
8
100KB < N ≤ 1MB
32
1MB < N ≤ 5MB
128
For example, if the message receiving volume is 5,000 messages per second and the message sending volume is 5,000 messages per second, with an average message body size of 8 KB, the message sending and receiving TPS is (8 / 4) x (5,000 + 5,000) = 20,000 times/second.

Pricing

Pricing of Cluster Specifications

TDMQ for Apache Pulsar pro clusters are divided into multiple tiers based on the total message sending and receiving TPS and peak bandwidth. Different tiers support varying computational workloads.
Item
Description
Billable item
Fees are charged based on the specifications of the cluster you purchased.
Billing mode
Billing cycle: determined by the actual duration you purchased.
Billing formula
Yearly/monthly subscription: Cluster specifications fee = Purchased duration (months) x Unit price (USD/month)
Billing unit price
For specific unit prices, see the actual sales page.

Pricing of Fixed Storage

TDMQ for Apache Pulsar pro clusters (fixed storage) allocate storage space dedicated to a specific cluster instance. Cloud SSD is provided as cluster storage with 3 replicas by default. You can purchase storage space on demand during cluster creation, with billing calculated linearly based on the purchased storage capacity and duration.
Storage specification fee = Selected specification (GB) x 3 x Purchased duration (months) x Unit price (USD/GB/month)
Region
Cloud Disk Type
South China (Guangzhou)
0.1550
East China (Shanghai)
0.1550
East China (Nanjing)
0.1550
North China (Beijing)
0.1550
Southwest China (Chengdu)
0.1550
Southwest China (Chongqing)
0.1550
East China (Shanghai Finance Zone)
0.2728
South China (Shenzhen Finance Zone)
0.2728
North China (Beijing Finance Zone)
0.2728
Hong Kong (China), Macao (China), and Taiwan (China)
0.1705
Southeast Asia Pacific (Singapore)
0.1938
Southeast Asia Pacific (Jakarta)
0.1938
Western United States (Silicon Valley)
0.1783
Europe (Frankfurt)
0.1938
Northeast Asia Pacific (Seoul)
0.2015
Eastern United States (Virginia)
0.1938
Southeast Asia Pacific (Bangkok)
0.1938
Northeast Asia Pacific (Tokyo)
0.2325
South America (São Paulo)
0.1938

Pricing of Elastic Storage

TDMQ for Apache Pulsar pro clusters (elastic storage) implement linear pricing for storage. Storage fee = Message storage size x Storage unit price x Duration.
Note:
Pulsar stores a message in three replicas by default. Therefore, the billable storage size is approximately three times the total message size.
The following table shows the unit price of message storage (Unit: USD/GB/hour).
Region
Guangzhou, Shanghai, Nanjing, Beijing, Chengdu, Chongqing, and Qingyuan
Shenzhen Finance, Shanghai Finance, and Beijing Finance
Hong Kong (China)
São Paulo, Singapore, Thailand, Jakarta, Virginia, and Frankfurt
Seoul and Northeast Europe
Japan
Silicon Valley
Unit price (USD/GB/hour)
0.00032551
0.00058902
0.00038751
0.00046502
0.00048052
0.00054252
0.00043401
Note:
1. Billing is calculated on an hourly basis, with the peak storage usage within each hour counted as the actual consumption. Any duration less than one hour is billed as a full hour.
2. Settlement is performed hourly, bills are pushed every 24 hours, and fees are automatically deducted from your account balance.

Pricing of Cross-Region Message-Level Replication

When you enable message-level cross-region replication tasks, fees are incurred, while metadata-level tasks do not incur fees. The console provides relevant monitoring metrics to facilitate observation.

Billable Items

Billable Item
Calculation Formula
Billing Method
Settlement Method
Message transfer
Message transfer volume (per 10,000 messages) x Unit price of message transfer
Postpaid
Settled hourly based on the cumulative value of messages. Bills are pushed daily.
Task resource usage
Number of topics in a task x Unit price of task resources
Postpaid
Settled daily based on the daily peak value, with usage under one day billed as a full day. Bills are pushed daily.

Price

Message Transfer
Number
Number of messages (100 million messages/month)
Unit price by region (USD/10,000 messages)
Guangzhou, Shanghai, Nanjing, Beijing, Chengdu, Chongqing, and Qingyuan
Financial Zone and Autonomous Driving Zone
Hong Kong (China), São Paulo, Singapore, Thailand, Jakarta, Virginia, Frankfurt, Seoul, Northeast Europe, Japan, and Silicon Valley
1
N < 1
0.0138
0.0221
0.0207
2
1 ≤ N < 10
0.0120
0.0199
0.0187
3
10 ≤ N
0.0111
0.0177
0.0166
Task Resource Usage
Unit price by region (USD/topic/day)
Guangzhou, Shanghai, Nanjing, Beijing, Chengdu, Chongqing, and Qingyuan
Financial Zone and Autonomous Driving Zone
Hong Kong (China), São Paulo, Singapore, Thailand, Jakarta, Virginia, Frankfurt, Seoul, Northeast Europe, Japan, and Silicon Valley
0.2766
0.4426
0.4149

Was this page helpful?
You can also Contact Sales or Submit a Ticket for help.
Yes
No

Feedback