This update covers features we are releasing this year near the Ignite 2023 conference for the Azure Service Bus, Azure Event Hubs, Azure Event Grid and Azure Stream Analytics services.
Azure Event Grid
MQTT broker
Reflecting the growing demand for connectivity, integration, and analytics between Internet of Things (IoT) devices and cloud-based services, Azure Event Grid's new MQTT broker feature is now Generally Available. Support for MQTT v5 protocol and MQTT v3.1. protocol enables publish subscribe over custom hierarchical topics. To process the data further, users can route IoT data to Azure services and 3rd party services using webhooks.
Flexible consumption of events
To enable event driven architectures in highly secure environments, pull delivery through namespace topics in Event Grid is now Generally Available. This allows customers to control the rate and volume of messages consumed and process events without configuring a public endpoint, while supporting much larger throughput.
Event Grid namespace now also supports the ability to push events to Azure Event Hubs at high scale through a namespace topic subscription. This is available in public preview and enables the development of more distributed applications to send discrete events to ingestion pipelines.
High scale
Each Namespace now enables connecting to 400,000 MQTT clients with up to 40MB per second or 40,000 MQTT messages per second ingress and egress; and 40MB per second or 40,000 events per second ingress, 80MB per second or 80,000 events per second egress.
To learn about new features released in Azure Event Grid, see this announcement.
Service Bus
Azure Service Bus partitioned namespaces GA
Partitioned namespaces, available in the Premium SKU for Azure Service Bus, enable our customers to run larger Service Bus workloads. With partitioned namespaces, your overall throughput is no longer limited to the performance of a single message broker. Using this new feature, messages are by default distributed across multiple brokers for a single queue or topic. This helps in multiple areas including increased throughput volume, improved reliability, and more. The new partitioned namespaces feature is also tied to a number of infrastructure changes, enabling availability zones everywhere where these are supported in Azure. To learn all about this great new feature, start with the announcement blog here:
Azure Event Hubs
Kafka Compression Public Preview
Azure Event Hubs now has support for Kafka compression. With compression enabled, your messaging workloads will take up less bandwidth, require fewer resources from the message brokers, and enable you to support larger uncompressed messages. This feature is now in public preview for Azure Event Hubs and is supported only for Apache Kafka traffic. There is more information at the announcement blog
Azure Stream Analytics
Kafka Input and Output (Public Preview)
Azure Stream Analytics now enables you to connect directly to Kafka clusters to ingest and output data with low code and high scalability. This solution is fully managed by the Azure Stream Analytics, ensuring compliance with business standards. The Kafka Adapters are backward compatible and support all Kafka versions from 0.10 onwards with the latest client release. You can connect to Kafka clusters inside a VNET or with a public endpoint, depending on your needs. The feature follows the existing Kafka configuration conventions and supports various compression types, such as None, Gzip, Snappy, LZ4, and Zst. To learn more, visit: https://aka.ms/asakafkaignite
Protobuf deserializer (Public Preview)
Azure Stream Analytics now supports a built-in protobuf deserializer that simplifies the ingestion of data as a protocol buffer. You can easily specify the file format, the protobuf definition file, the message type and the prefix style for your input data. This solution offers a more user-friendly experience than using a custom deserializer. To learn more, visit: https://aka.ms/asaprotobufdeserializer