GithubHelp home page GithubHelp logo

azure-relay-java's Introduction



Microsoft Azure Relay

Microsoft Azure Relay Hybrid Connections for Java

This library is built using Java 8. Please ensure that you have JDK 1.8+ installed before running this project.

Azure Relay is one of the key capability pillars of the Azure Service Bus platform. The Relay’s "Hybrid Connections" capability is a secure, open-protocol evolution based on HTTP and WebSockets. It supersedes the former, equally named "BizTalk Services" feature that was built on a proprietary protocol foundation. The integration of Hybrid Connections into Azure App Services will continue to function as-is.

"Hybrid Connections" allows establishing bi-directional, binary stream communication between two networked applications, whereby either or both parties can reside behind NATs or Firewalls. This document describes the client-side interactions with the Hybrid Connections relay for connecting clients in listener and sender roles and how listeners accept new connections.

How to provide feedback

See our Contribution Guidelines.

Samples

For Relay Hybrid Connections samples for Java, see the azure/azure-relay service repository.

Registering Hybrid Connections

To use the Hybrid Connection feature, you must first register a Hybrid Connection path with the Relay service. Hybrid Connection paths are string expressions that uniquely identify the entity.

To create Hybrid Connection entities, you first need a Service Bus Relay namespace that you can create either through the Azure portal, the Azure PowerShell tools, or the cross-platform Azure CLI.

Existing Azure Relay namespaces can be managed in the Azure Portal, where you can also add, edit, or remove Hybrid Connection paths interactively.

The following two settings are specific to Hybrid Connections:

Property Description
RequiresClientAuthorization If this is set to false (the default is true), sending clients can connect to a listener through the Relay without providing an authorization token. In this case, the Relay will not enforce any if its ownaccess rules, but the listener can still evaluate the Authorization HTTP header or use some other model for access control.
ListenerCount This is an informational value that’s available via GetRuntimeInformationAsync and gives the number of connected listeners on this Hybrid Connection as the value is queried.

Up to 25 listeners can be concurrently connected and the Relay will distribute incoming connection requests across all connected listeners, equivalent to a network load balancer.

Handling Tokens

Creating a listener requires an access token that confers the "Listen" right on the Hybrid Connection entity or at the namespace level. Creating a sender connection requires, unless the Hybrid Connection entity is configured otherwise, a token that confers the "Send" right. The follows the shared access signature authentication model that is common across all Service Bus capabilities and entities.

Access tokens are created from an Authorization rule and key using a token provider helper as described in the article linked above; the Hybrid Connections API has its own TokenProvider class, however. The TokenProvider can be initialized from a rule and key with TokenProvider.CreateSharedAccessSignatureTokenProvider(ruleName, key) or it can be initialized from an existing token string that has been issued by some other application with TokenProvider.CreateSharedAccessSignatureTokenProvider(token).

The initialized TokenProvider instance is used by the HybridConnectionListener and HybridConnectionClient API to create tokens as needed.

However, with Hybrid Connections even more than with other Service Bus features, you may have scenarios where you will want the Relay to protect your endpoint, but you also don’t want to hand the SAS rule and key to the client outright. One such case are browser-based clients. For a browser-based client that needs to connect to a resource made available via a relayed WebSocket, the server-side web site can hold on to the required SAS rule and key, and use the TokenProvider to create a short-lived token string and pass that on to the client:

TokenProvider tokenProvider = TokenProvider.createSharedAccessSignatureTokenProvider(SharedAccessKeyName, SharedAccessKey());
String tokenString = tokenProvider.getTokenAsync("http://namespace.servicebus.windows.net/path", Duration.ofSeconds(30)).join().getToken();

The token created in the exemplary snippet above will only be valid to establish a connection within 30 seconds of receiving it.

Creating Listeners

The Hybrid Connection API follows a very common networking design pattern. There is a listener object that is first opened to allow incoming connections to flow and from which the application can then accept these incoming connections for handling.

With the listener object successfully instantiated, it should connect to the Azure Relay cloud service and listen for any incoming connections using listener.openAsync(timeout). After having connecting to the Azure Relay cloud service, the listener will be able to accept HTTP or web socket connections, then receive and send messages from these connections. The example illustrates some of the typical usages of the listener object, and more detailed samples may be found here: Azure Relay Java Examples.

HybridConnectionListener listener = new HybridConnectionListener("sb://namespace.servicebus.windows.net/path", tokenProvider); 
listener.openAsync(Duration.ofSeconds(60)).join(); 

listener.setRequestHandler((context) -> {
	// Handle HTTP requests in this handler

	RelayedHttpListenerResponse response = context.getResponse();
	response.setStatusCode(202);
	response.setStatusDescription("OK");
	resonse.close();
});

do 
{
	// This accepts connections with remote web sockets and provides a connected web socket instance
	ClientWebSocket websocket = listener.acceptConnectionAsync().join(); 
	
	// Send and receive messages with the connected websocket object
} 
while( … );

The HybridConnectionListener will aggressively attempt to stay connected once opened. Should the local network connection drop or connectivity to the Relay become interrupted, the listener will patiently retry until the listener can be restored.

ClientWebSocket is a web socket implementation based on javax.websocket from the Jetty library, and it will support many asynchronous operations such as sendAsync(), receiveMessageAsync() and closeAsync(). The usage of CompletableFuture within the implementation provides many useful hooks to insert handlers for different events.

Creating Clients

Client connections are created using the HybridConnectionClient class. There are two variants of the constructor: one takes the target address and a TokenProvider that can produce a "Send" token for the target. The other omits the token provider for use with Hybrid Connections that are set up without client authorization.

New web socket connections are created via the createConnectionAsync() method. When the connection has been established, the CompletableFuture completes with a ClientWebSocket that is connected to the remote listener. If the connection attempt fails, a RuntimeException will be raised that indicates the reason for why the connection could not be established.

How do I run the unit tests?

The unit tests are written using JUnit. You can simply run the unit tests from within an IDE such as Eclipse or Intellij by using their respective interfaces.

If running from within command prompt, simply navigate to the root folder of the project (/azure-relay-java) and run mvn clean test.

azure-relay-java's People

Contributors

bainian12345 avatar dlstucki avatar microsoftopensource avatar msftgits avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.