In this episode, Clemens introduces CNCF xRegistry (xregistry.io), explains how it evolved from CloudEvents work in the CNCF Serverless Working Group, and shows why teams need machine-readable metadata for events, schemas, and endpoints to build reliable event-driven systems. He walks through the core idea of an extensible, model-driven registry with consistent APIs, built-in versioning, and support for common formats and protocols, then highlights the surrounding tooling (server, CLI, code generator, and viewer). The episode closes with a practical look at how Microsoft Fabric uses xRegistry concepts to power typed event pipelines and schema-aware integration, and previews deeper dives in upcoming sessions.

https://cloudevents.io

In this episode of Subscribe, Clemens discusses CNCF CloudEvents, a protocol suite developed by the CNCF Serverless Working Group to provide a common metadata language for events. Developed by a collaborative group of up to 40 companies, CloudEvents establishes a layered standard that defines what an event is and provides consistent mechanisms for expressing those events across various applications, encodings, and transport options such as Kafka, HTTP, and MQTT. The standard addresses the challenges of modern, multi-hop architectures by ensuring that critical contextual information is retained as events move through IoT gateways, brokers, and analytics engines, thereby enabling infrastructure to effectively filter and route event traffic.

In this episode of Subscribe, Clemens introduces JSON Structure, a pragmatic data definition language designed to address the architectural shortcomings of JSON Schema. Consisting of a core specification and six companion extensions, the project shifts focus from simple validation to robust structural modeling for programming languages and databases. Key features include a rich system of extended primitive types like 128-bit integers and UUIDs, first-class compound types such as sets, maps, and tuples, and a disciplined approach to type reuse through namespaces and strict $ref rules. By decoupling structural definition from pattern-based validation and introducing an explicit import model, JSON Structure provides a deterministic, tooling-friendly alternative for high-interoperability data interchange

In this episode, the relationship and shared architecture between Azure Stream Analytics (ASA) and Microsoft Fabric Event Streams are explored, highlighting how both leverage the same underlying engine to process real-time data with millisecond latency. The session details how ASA functions as an event-driven engine for creating standing queries using a SQL-92 dialect or a no-code drag-and-drop interface to yield insights, aggregations, and enrichments on the "hot path" before data hits a database. It is further explained that Fabric Event Streams provides an opinionated, integrated experience that combines ASA and Event Hubs, offering an expanded set of connectors for CDC, external Kafka, and cloud platforms like AWS and GCP. The episode concludes by emphasizing that while Fabric simplifies the user experience, it retains the full power of ASA’s temporal windowing and developer tooling compatibility, making it a high-throughput foundation for modern real-time analytics.

In this fifth episode, we explore Azure Event Grid, the versatile eventing backplane for Microsoft Azure and Microsoft Fabric. This episode details how Event Grid functions as a unified system combining a fully conformant MQTT 3.1.1 and 5.0 broker with a CloudEvents-based PubSub engine. We break down the technical differences between the Basic tier, which focuses on event-driven push delivery to targets like Azure Functions and Webhooks, and the Standard tier, which introduces namespaces, VNet support, and queue-like pull delivery. Whether you are managing IoT telemetry from hundreds of thousands of devices or building serverless architectures, you will learn how Event Grid handles complex routing, retry logic, and integration with Microsoft Fabric.

In this episode, Clemens Vasters explains that Azure Service Bus is a fully managed message queue and PubSub broker that serves as a fundamental infrastructure piece for distributed systems. He defines the service as reliability as a service because it provides individual state management for the lifecycle of every message, ensuring that messages are exclusively acquired by a single consumer to prevent duplicate processing. The system supports industry-leading reliability by writing messages to three replicas across zones or using geo-replication across regions to ensure data is acknowledged only after it is safely stored. Vasters notes that the service is AMQP 1.0 standards compliant and is the only hyperscale broker fully conformant with JMS 2.0, making it a highly compatible choice for Java applications and various SDKs.The talk highlights the architectural differences between the Standard and Premium SKUs, noting that Premium offers isolated CPU and memory to eliminate noisy neighbor problems found in shared infrastructure. Vasters describes how topics and subscriptions allow for a distribution mechanism where messages sent to a single address can be fanned out to multiple parties using SQL-92 based filters and actions. Beyond basic messaging, the service acts as a secure layer 7 application router that can bind to multiple virtual networks and public endpoints simultaneously. This configuration allows messages to pass across network boundaries without the networks actually touching, providing a high level of isolation and security for production workloads.

In this episode, Clemens introduces Azure Event Hubs as Azure’s high‑throughput event streaming backbone, handling tens of trillions of events per day and exposing a fully managed, standards-based event stream broker that supports AMQP, HTTP, and the Apache Kafka protocol plus SDKs for major languages. He explains the different SKUs—Standard (shared, low-cost), Premium (isolated resources and advanced networking), and Dedicated (for very large sustained workloads)—and describes the architecture of namespaces, event hubs (topics), partitions as replicated logs, and consumer groups that coordinate readers over the log. Event Hubs is positioned not as a queue, database, or generic pub/sub broker, but as a fast, durable buffer for time-ordered event ingestion and replay with configurable retention, with reliability provided by zone-replicated partitions, optional cross-region geo-replication (async or synchronous with RPO‑0), strong virtualized networking and private endpoints, and Event Hubs Capture, which automatically archives incoming events into Avro files in Azure Storage for long-term retention and batch processing.

This episode of Subscribe provides a 10 minute overview of Microsoft Fabric Real-Time Intelligence, a unified solution designed to close the digital feedback loop by capturing and acting on data at the speed of business. While traditional real-time pipelines are often cost-prohibitive due to complex integration requirements, this platform simplifies the process for data analysts by layering an opinionated application on top of battle-hardened infrastructure like Event Hubs, Stream Analytics, and Kusto (KQL) Database. We explore how the Real-Time Hub acts as a centralized catalog for managing diverse data streams, from Kafka and AMQP endpoints to Change Data Capture (CDC) from SQL databases, all while maintaining automated schema propagation.