tencent cloud

TDMQ for Apache Pulsar is Tencent's proprietary messaging middleware based on Apache Pulsar. It comes with excellent cloud native and serverless features and is compatible with all components and principles of Pulsar. It also has the underlying benefits of computing-storage separation and flexible scaling.

Strong Data Consistency

By writing message data backups to various physical machines with sync flush, TDMQ for Apache Pulsar achieves strong data consistency by using BookKeeper consistency protocol. When one of the physical machines fails, the backend data replication mechanism can quickly migrate the data to guarantee data backups are available.

High Performance and Low Latency

With over 100,000 QPS per cluster, TDMQ for Apache Pulsar can easily maintain the production and consumption of millions of messages, as well as retain a massive number of messages. It also offers a duration protection mechanism to ensure minimal latency and help you easily meet business performance requirements.

Millions of Topics

TDMQ for Apache Pulsar's computing and storage structures are designed to be independent of one another, allowing it to support millions of message topics with ease. When compared to other message queue products on the market, the performance of a TDMQ for Apache Pulsar cluster will not suffer much as the number of topics increases.

Rich Diversity of Message Types

TDMQ for Apache Pulsar offers a rich diversity of message types, such as general, sequential (global and partitioned), distributed transaction, and scheduled messages, meeting the requirements for advanced features in various demanding scenarios.

Unlimited Consumers

The number of consumers is not limited by the number of topics in TDMQ for Apache Pulsar, and the quantity of messages per consumer is balanced using algorithms. Businesses can start with the appropriate number of consumers as needed.

Isolation Control

TDMQ for Apache Pulsar offers a mechanism of topic isolation by tenant. It accurately controls the production and consumption speeds of each tenant, prevents the tenants from affecting each other, and ensures that message processing won't cause resource competition.


The data of each transaction order needs to be monitored by dozens of downstream business systems, including item price approval, delivery, reward point, and stream computing analysis. Such systems use different message processing logic, making it impossible for a single system to adapt to all associated business. In this case, TDMQ for Apache Pulsar can implement efficient async communication and application decoupling to ensure the business continuity of the primary site.

Companies hold promotional campaigns such as new product launch and festival red packet grabbing from time to time, which often cause temporary traffic spikes and pose huge challenges to each backend application system. In this case, TDMQ for Apache Pulsar can act as a buffer to centrally collect the suddenly increased requests in the upstream, allowing downstream businesses to consume the request messages based on their actual processing capacities.

Sequential messages are used in some business scenarios, such as order creation, payment, delivery, and refund of in-app/game items, which are all strictly executed in sequence. Similar to the First In, First Out (FIFO) principle, TDMQ for Apache Pulsar offers a sequential message feature dedicated to such scenarios to ensure message FIFO.

TDMQ for Apache Pulsar can easily implement cross-IDC sync if messages need to be consumed across many IDCs.

Contact Us

Click "Contact Us" to enjoy personalized presales consultation and aftersales service.