Table of Contents
What is a message broker and what are its benefits
A message broker is tool that queues and processes messages received from the end-point consumers. Message brokers can be used for many purposes: message routing, decoupling application responses from their consumers, isolating failures in service teams. A message broker can be an important element of enterprise SOA toolkit, as it decouples the messaging and processing logic.
RESTful web service message broker based on the JRocket framework.
##
Message brokers [link] : http://en.wikipedia.org/wiki/Message_broker | http://en.wikipedia.org/wiki/Brokerage
messaging [link] : http://en.wikipedia.org/wiki/Messaging | http://www.purdueglobalacademy.com/wp-content/uploads/2013/10/Messaging-10122013-1schallenge_slides%202.pdf
##
The different types of message brokers available
Message brokers can be classified into three different types: message queuing, service-oriented message brokers, and messaging as a service. Message queuing – When only a small number of message producers and consumers send and receive messages. [Example: Exchange Server] Service-oriented message brokers – A service-oriented message broker can support a richer set of functionalities, such as retry logic, message queues, event management and internal messaging. [Example: WCF Message Broker] Messaging as a service – Message brokers that are available as public services. In this case one does not have to install the actual broker software on the application server. [Example: IBM MQ Series]
In the following, we will take a closer look at the three different message broker types mentioned above to understand the functionalities that they offer and how they compare to each other. They can be considered as follows [Example: Messaging node in web sphere]:
How to choose the right message broker for your needs
The first step in choosing a message broker is to identify your use cases, the messages that need to be exchanged, how many subscribers are involved and how frequently messages are being sent. Once these questions are answered, you can decide whether you need a message queuing or service-oriented model. If messaging needs to scale up or down depending on the number of subscribers, then service-oriented implementation is the best choice. If the publisher needs to have control over internal messaging, then you should consider message queuing. If scalability for many subscribers is not a priority, or you want to decouple the application logic from the messages being delivered, then a service-oriented message broker could be a better choice. The following is a summary of the different service-oriented message broker models and their benefits and drawbacks.
##Example of the deployment of a message queue:##
##Example of the deployment of service-oriented message broker:##
In both cases, more advanced features such as support for transactional messaging or role-based access control can be added by wrapping the broker with another application, but we will not discuss those topics in this article.
##Example of the deployment of a message broker supporting transactional messaging:##
It’s important to note that even though service-oriented solutions are good for decoupling application logic from the messages they send, they might not be the best solution for all applications. In some cases, it might make sense to combine a message broker with another messaging system – especially if your application supports sending messages to multiple subscribers.
To summarize our discussion on message brokers, we recommend service-oriented solutions when you need to decouple your application logic from the messaging system and a queuing model when you need to scale up or down the system at runtime.
__Example of the deployment of a message broker supporting both services and queues:__
The features you should look for in a message broker
A message broker should have the ability to prescreen incoming messages, accept duplex messages and support serialization. It must be able to monitor the queue for new messages, react to various kinds of exceptions or failures and replay failed messages. Message brokers with sophisticated message routing ability can support several routing models and can also be used for internal messaging. Lastly, it must be fault tolerant and its operation should be transparent to the system.
A message broker is deemed a central nervous system for applications and middleware. It is essential for an enterprise to choose the right message broker that will meet its requirements. To help you make an informed decision, here is a simplified rundown of the features that you should look for in a message broker.
Prescreening of incoming messages
Prescreening incoming messages may be necessary if your application has stringent security or performance requirements. A common use of this feature is to filter out messages with invalid or unauthorized source addresses.
When you receive a message from an unknown source, you may want to investigate before accepting the message into your application. Prescreening can also be used to enforce security policies about who should get a certain set of messages, based on membership in a given list.
Accepting duplex messages
How to get started using a message broker
Any message broker can be implemented in Visual Basic or C# and there are a number of open source message brokers such as Apache ActiveMQ, JMS/JGroups, Aquto and others. If you’re not familiar with programming languages or operating systems, then it’s also possible to use web services for the implementation using existing web services frameworks. Since message brokers are usually implemented as web services, it’s easy to implement any kind of web service around one.
The basic concept of the message broker is to treat each part of a distributed system as one process or one program. This might seem like a simple thing, but this is where it gets interesting because if you’re working with multiple computers, each computer will require its own set of hardware. It’s also possible for a computer to be running several applications on top of being a client or server for other applications.
