According to news from IT House on July 14, Baidu yesterday open sourced its MQTT Broker message middleware BifroMQ, which is implemented by Java and is characterized by "high performance and distributed" . The basic technology of the cloud IoT core suite IoT Core".
According to reports, BifroMQ adopts the Serverless architecture and seamlessly integrates native multi-tenant support . This message middleware is derived from the Baidu IoT team's years of technology accumulation, and aims to support the construction of large-scale IoT device connections and message systems .
BifroMQ uses a load-independent sub-cluster design that can " efficiently handle workloads such as connection sessions, message forwarding, and message storage." Each sub-cluster has "high availability, supports horizontal expansion, and can flexibly respond to multi-tenant scenarios ".
At the same time, BifroMQ has a built-in optimized distributed storage engine, which can ensure stable performance without relying on third-party storage middleware .
In addition, BifroMQ supports three deployment modes: stand-alone mode Standalone, standard cluster Standard Cluster, and independent workload cluster Independent Workload Cluster , and allows customization through plug-ins to meet different business needs.
In terms of performance, in the standard test environment, when dealing with a large number of concurrent message publishing, BifroMQ maintains low message latency and low CPU usage.
In terms of usage scenarios, Baidu claims that BifroMQ can be widely used in various IoT scenarios, including but not limited to smart homes, industrial IoT, Internet of Vehicles, and smart cities. BifroMQ can provide reliable and high-performance message transmission services for these applications, supporting large-scale device connections and message processing.
Baidu said that BifroMQ is a key open source project. BifroMQ will continue to improve its support for MQTT 5 in the future. They will invest more resources to support the development of BifroMQ to meet more business needs.
Deploy with Docker
Deploy on Linux or macOS
Deploy on Windows
Compile and deploy from source
Please see the detailed steps in the comment area