Baidu Messaging System (BMS)
- Overview
- Features
- Advantages
- User’ s Guide
Overview
BMS is a distributed and scalable hosting message queue service with high throughput, and it supports the following business scenarios: · It collects massive data from websites, devices, or applications for the real-time analysis, such as user browsing, click, and search. · It gathers the remote sensing data of distributed application to facilitate the system operations. · It interfaces with the Spark Streaming and other services for the real-time stream data analysis. BMS is a hosted service based on Apache Kafka. Kafka is a distributed, multi-partition and multi-replica messaging service. The producer asynchronously interacts with the consumer through the message queue without waiting for each other. Compared with the traditional messaging service, Kafka features as follows: · The topic can be scaled horizontally through the partition. · The partition is distributed on multiple nodes to achieve the high availability of data. · It supports the individual consumer to consume messages in the form of queue or Pub/Sub or multiple consumers to consume messages in the cluster order through the consumer group. BMS encapsulates the Kafka cluster details, and provides them in the form of hosted service. You can directly use BMS to integrate with the massive distributed applications without the consideration of cluster operations and pay per use only.
Features
It is fully compatible with the open-source community Kafka. After the service is activated, you can create the Kafka topic, and use the Kafka client to communicate with BMS.
You can continuously collect the massive data to respond to the event, equipment and business real time.
BMS can scale up on an on-demand basis. Compared with the traditional hardware for which the budget plan is made and that is purchased based on maximum usage, it is convenient and paid as per the usage of messages only.
BMS provides SLA to ensure availability. Furthermore, it provides multiple replicas of the data to ensure the persistence.
Advantages
It is activated to use the Kafka system. It focuses on product development without considering the installation, deployment, configuration, debugging, and maintenance of clusters.
You do not need to invest in any hardware and software, but activate the service and pay per the resources used. For it is compatible with the community Kafka, the migration cost is extremely low, and you do not need to worry about being locked by the technology.
BMS supports the SSL encrypted data transmission only to ensure that the data is not eavesdropped or tampered during the transmission, and ensure the security of customer data.
The unique high-availability design can prevent data loss in case of application failure, individual machine failure or facility failure, and keep the data for 7 days at most.