Why Huawei Cloud DMS?
Rapid Deployment, Open-source Compatibility
Works with mainstream message protocols, such as Kafka, RocketMQ, and RabbitMQ (AMQP)
Usable right out of the box and managed and integrated with RESTful APIs, console, and multi-language SDKs
24/7 O&M Assurance, Monitoring, Fault Detection
Guard reliability with multi-replica, multi-AZ, and failover
Stay informed with one-click diagnosis (DMS for RocketMQ), comprehensive monitoring, and service assurance around the clock
.png)
TPS in the Tens of Millions
Real-time data processing with performance at 200 Gbit/s and latency of seconds
Storage and computing resources adjustable to resource challenges in different scenarios
High-Throughput Message Queues for All Needs High-Throughput Message Queues for All Needs
Asynchronous communication
Sub-systems decoupled for a monolithic application to become distributed, with higher stability and performance
IoT
High-speed transmission channels between IoT devices and data analytics systems
Stream and big data processing
A streaming model that captures and processes data surges in real time
Data transfer hub
Ultra-high-capacity storage, one-to-many consumption, and real-time and batch processing of data between systems
Asynchronous communication
Sub-systems decoupled for a monolithic application to become distributed, with higher stability and performance
IoT
High-speed transmission channels between IoT devices and data analytics systems
Stream and big data processing
A streaming model that captures and processes data surges in real time
Data transfer hub
Ultra-high-capacity storage, one-to-many consumption, and real-time and batch processing of data between systems
DMS Feature Differences
DMS Feature Differences
Feature
DMS for Kafka
DMS for RocketMQ
DMS for RabbitMQ
Recommended setting: 0–10
Pull-based
Pull-based and push-based
Pull-based and push-based
Offset and timestamp tracking
Once a message retrieval has been acknowledged, RabbitMQ will be notified that the message can be deleted
Higher accumulation performance than RabbitMQ thanks to high throughput
Supported by the firehose feature or the rabbitmq_tracing plugin. However, rabbitmq_tracing reduces performance and should be used only for troubleshooting
Not supported, but can be encapsulated
Only supports Apache Kafka
Compatible with RocketMQ
RabbitMQ is based on AMQP
Kafka is written in Scala and Java and supports clients in multiple programming languages
Supports clients in multiple programming languages
RabbitMQ is written in Erlang and supports clients in multiple programming languages
Throttling on users and producer or consumer clients
Planned
Credit-based throttling on producers, a mechanism that triggers protection from within
Supports partition-level FIFO
Message order is maintained within a queue
FIFO only for single-threaded message queuing without advanced features such as delayed queues or priority queues
SSL and SASL authentication and read/write permissions control
SSL authentication
SSL authentication
DMS for Kafka
Pull-based
Offset and timestamp tracking
Higher accumulation performance than RabbitMQ thanks to high throughput
Only supports Apache Kafka
Kafka is written in Scala and Java and supports clients in multiple programming languages
Throttling on users and producer or consumer clients
Supports partition-level FIFO
SSL and SASL authentication and read/write permissions control
DMS for RocketMQ
Pull-based and push-based
Compatible with RocketMQ
Supports clients in multiple programming languages
Planned
Message order is maintained within a queue
SSL authentication
DMS for RabbitMQ
Recommended setting: 0–10
Pull-based and push-based
Once a message retrieval has been acknowledged, RabbitMQ will be notified that the message can be deleted
Supported by the firehose feature or the rabbitmq_tracing plugin. However, rabbitmq_tracing reduces performance and should be used only for troubleshooting
Not supported, but can be encapsulated
RabbitMQ is based on AMQP
RabbitMQ is written in Erlang and supports clients in multiple programming languages
Credit-based throttling on producers, a mechanism that triggers protection from within
FIFO only for single-threaded message queuing without advanced features such as delayed queues or priority queues
SSL authentication
Documentation
More Services
More Services
FunctionGraph
Run your code without provisioning or managing servers
Distributed Cache Service (DCS) for Redis
An in-memory caching service compatible with Redis
Cloud Service Engine (CSE)
Cover microservice application registration, discovery, configuration management, and governance