Big data is becoming more and more important, and more and more products rely on big data computing. If your product is deployed in the user room, edge end, or private cloud, when the product is exported, a series of big data clusters will be deployed, such as: Collecting cluster, stream computing cluster, message queue cluster, each cluster at least three machines to guarantee the reliability of the cluster, even if there is mixed, output resources is also very big, especially the flow calculation need to task set ahead of resources, and each task resources are also expensive, insufficient resources quickly, also need expansion (purchase, approval, deployment). RocketMQ Streams integrates collection capabilities, message queues, and stream computing capabilities. It can be deployed with 1core and 1 gigabyte, and can be scaled up by adding instances as web applications do, especially for highly filtered scenarios. Resources do not increase linearly with rules, and CPU resources are 40% of those of similar big data computing. The CPU is 4%.

In this topic, you can not only understand the implementation principle of RocketMQ-Streams, but also determine whether your business can have a lighter solution, lower product cost and provide competitiveness based on ali’s rocketMQ-Streams application practice

Speakers:

Xiaodong Yuan: Ali cloud, Senior technical Specialist, Yuan Xiaodong: Apache RocketMQ Committer, RocketMQ-Streams Cofounder, Ali Cloud senior technical expert, responsible person of security intelligent computing engine