Event Grid on Kubernetes - Concepts

This article describes the main concepts in Event Grid on Kubernetes with Azure Arc (Preview).

Important

Event Grid on Kubernetes with Azure Arc is currently in public preview. This preview version is provided without a service level agreement, and it's not recommended for production workloads. Certain features might not be supported or might have constrained capabilities. For more information, see Supplemental Terms of Use for Microsoft Azure Previews.

Events

An event is a data record that announces a fact about the operation of a software system. Typically, an event announces a state change because of a signal raised by the system or a signal observed by the system. Events contain two types of information:

• Event data that represents the occurrence of a state change.

• Context attributes that provide contextual information about the occurrence of the event.

  Both event data and context attributes can be used for filtering events.

Event Grid on Kubernetes supports CloudEvents schema specification. Here's an example of an event that uses the CloudEvents schema. Event Grid supports an event of size up to 1 MB.

[{
       "specVersion": "1.0",
       "type" : "orderCreated",
       "source": "myCompanyName/us/webCommerceChannel/myOnlineCommerceSiteBrandName",
       "id" : "eventId-n",
       "time" : "2020-12-25T20:54:07+00:00",
       "subject" : "account/acct-123224/order/o-123456",
       "dataSchema" : "1.0",
       "data" : {
          "orderId" : "123",
          "orderType" : "PO",
          "reference" : "https://www.myCompanyName.com/orders/123"
      }
}]

Source

The source attribute describes the context in which the event occurred. The source may be the originator of events. However, in some cases, there are producers that create and publish events. And, these producers are distinct from the source. For simplicity, this article assumes that the source is the producer of the events.

Each event source produces events of one or more event types. As the source of events, an application defines a set of related events to announce its state changes. Every event has common information such as the source of the event, time the event took place, and a unique identifier. Every event also has specific information that's only relevant to the specific type of event. The support for an event of size up to 1 MB is currently in preview.

For the properties that are included in an event, see CloudEvents schema.

Publishers

Event publishers are applications or systems that send events to Event Grid to be delivered to event subscribers.

Topics

A topic is a form of input channel that exposes an endpoint to which publishers send events to Event Grid.

A topic can be used for a collection of related events. You could create a topic for each category of related events. In some cases, the source can be used to organize events into categories because sources are usually associated with a set of closely related event types (“MyApp.OrderCreated”, “MyApp.OderDeleted”, “MyApp.OrderRejected”, and so on).

Consider an application that sends events related to managing user accounts and processing orders. It's unlikely that an event subscriber is interested in consuming both categories of events. Create two custom topics and let event handlers subscribe to the one that interests them. For small solutions, you might prefer to send all events to a single topic.

Event subscribers

Event subscribers are software systems such as microservices that expose endpoints to which Event Grid delivers events.

Event subscriptions

An event subscription tells Event Grid which events on a topic you're interested in receiving (event filtering) and where to send them (event routing). When creating an event subscription, you provide an endpoint for handling the event. You can select the events that you want to be delivered to your endpoint by configuring filter clauses on the event subscription.

Event handlers

An event handler is a software system that exposes an endpoint to which events are sent. The handler receives the event and takes actions to process the event. Event Grid supports several handler types. As the handler, you can use a supported Azure service hosted on Kubernetes or Azure, or your own solution that exposes a web hook (endpoint) wherever that's hosted. Depending on the type of handler, Event Grid follows different mechanisms to guarantee the delivery of the event. If the destination event handler is an HTTP web hook, the event is retried until the handler returns a status code of 200 – OK. For more information, see Event handlers.

SAS authentication

Event Grid on Kubernetes provides SAS key-based authentication for publishing events to topics.

Event delivery

Event Grid on Kubernetes provides a reliable delivery and retry mechanism. If Event Grid can't confirm that an event has been received by the event handler endpoint, it redelivers the event. For more information, see Event Grid message delivery and retry.

Batch event publishing

When you use a topic, events must always be published in an array. For low throughput scenarios, the array will have only one event. For high volume use cases, we recommend that you batch several events together per publish to achieve higher efficiency. Batches can be up to 1 MB. Each event should still not be greater than 1 MB. For more information, see Batch event delivery.

Event Grid on Kubernetes components

• The Event Grid operator implements the Operator pattern. It watches for state changes of Event Grid resources as a result of control plane requests made to Kubernetes' API Server. When there's a request that affects the state of any of Event Grid resources, the Event Grid operator syncs that state with the Event Grid Broker.

• The Event Grid broker serves as both control plane and data plane operations.

  As a control plane service, it's responsible for bringing the state of Event Grid to the desired state communicated by the Event Grid Operator. For example, when a request is made to create a new topic, it fulfills that request and the service metadata is updated.

  As a data plane service, it serves all event publishing requests and delivers events to their destinations configured on event subscriptions.

Next steps

To get started, see Create topics and subscriptions.