The transaction engine is a core system for the billing of Tencent products. The order data of each transaction is used by dozens of downstream business systems, including the item wholesale price, prop delivery, points, and stream computing. Different systems have different message processing logic, and systems cannot adapt to each associated business. TDMQ for Apache Pulsar can achieve efficient asynchronous communication and application decoupling, ensuring primary site business continuity.
Peak Shifting
Enterprises hold marketing campaigns, new product launches, and festival red packet events at irregular intervals, which typically generate temporary traffic surges. These events impose significant pressure on backend application systems. Scale-out could address the issue, but may lead to resource waste. In such scenarios, TDMQ for Apache Pulsar serves as a buffer and centrally collects messages that increase suddenly from upstream systems. Downstream systems can consume request messages based on their actual processing capabilities.
Sequential Sending and Receiving
The sequential message sending and receiving feature is applied to business scenarios. For example, in Honor of Kings, the purchase and distribution of skin items involve processes, such as order creation, payment, and refund, which are strictly executed in sequence, similar to the First In First Out (FIFO) principle. TDMQ for Apache Pulsar provides the sequential message sending and receiving feature, specifically designed for such scenarios, ensuring FIFO of messages.
Data Synchronization
If there are multiple data centers and consumption is required across them, TDMQ for Apache Pulsar can conveniently facilitate synchronization between the data centers.
Big Data Analysis
Value is generated from data as it "flows". Traditional data analysis mostly relies on batch computing models, which cannot achieve real-time data analysis. Combining TDMQ for Apache Pulsar with stream processing engines enables convenient real-time analysis of business data.
