Integration

The default constructor uses java.net.HttpURLConnection to perform requests. You can switch to a different HTTP library with an implementation of ClientHttpRequestFactory. There is built-in support for the following:

For example, to switch to Apache HttpComponents, you can use the following:

Each ClientHttpRequestFactory exposes configuration options specific to the underlying HTTP client library — for example, for credentials, connection pooling, and other details.

Objects passed into and returned from RestTemplate methods are converted to and from raw content with the help of an HttpMessageConverter.

On a POST, an input object is serialized to the request body, as the following example shows:

You need not explicitly set the Content-Type header of the request. In most cases, you can find a compatible message converter based on the source Object type, and the chosen message converter sets the content type accordingly. If necessary, you can use the exchange methods to explicitly provide the Content-Type request header, and that, in turn, influences what message converter is selected.

On a GET, the body of the response is deserialized to an output Object, as the following example shows:

The Accept header of the request does not need to be explicitly set. In most cases, a compatible message converter can be found based on the expected response type, which then helps to populate the Accept header. If necessary, you can use the exchange methods to provide the Accept header explicitly.

By default, RestTemplate registers all built-in message converters, depending on classpath checks that help to determine what optional conversion libraries are present. You can also set the message converters to use explicitly.

WebFlux

The spring-web module contains the HttpMessageConverter contract for reading and writing the body of HTTP requests and responses through InputStream and OutputStream. HttpMessageConverter instances are used on the client side (for example, in the RestTemplate) and on the server side (for example, in Spring MVC REST controllers).

Concrete implementations for the main media (MIME) types are provided in the framework and are, by default, registered with the RestTemplate on the client side and with RequestMappingHandlerAdapter on the server side (see Configuring Message Converters).

The implementations of HttpMessageConverter are described in the following sections. For all converters, a default media type is used, but you can override it by setting the supportedMediaTypes bean property. The following table describes each implementation:

You can specify a Jackson JSON View to serialize only a subset of the object properties, as the following example shows:

To send multipart data, you need to provide a MultiValueMap<String, Object> whose values may be an Object for part content, a Resource for a file part, or an HttpEntity for part content with headers. For example:

In most cases, you do not have to specify the Content-Type for each part. The content type is determined automatically based on the HttpMessageConverter chosen to serialize it or, in the case of a Resource based on the file extension. If necessary, you can explicitly provide the MediaType with an HttpEntity wrapper.

Once the MultiValueMap is ready, you can pass it to the RestTemplate, as show below:

If the MultiValueMap contains at least one non-String value, the Content-Type is set to multipart/form-data by the FormHttpMessageConverter. If the MultiValueMap has String values the Content-Type is defaulted to application/x-www-form-urlencoded. If necessary the Content-Type may also be set explicitly.

The Spring Framework lets you define an HTTP service as a Java interface with annotated methods for HTTP exchanges. You can then generate a proxy that implements this interface and performs the exchanges. This helps to simplify HTTP remote access which often involves a facade that wraps the details of using the underlying HTTP client.

One, declare an interface with @HttpExchange methods:

Two, create a proxy that will perform the declared HTTP exchanges:

@HttpExchange is supported at the type level where it applies to all methods:

Annotated, HTTP exchange methods support flexible method signatures with the following method parameters:

Annotated, HTTP exchange methods support the following return values:

By default, WebClient raises WebClientResponseException for 4xx and 5xx HTTP status codes. To customize this, you can register a response status handler that applies to all responses performed through the client:

For more details and options, such as suppressing error status codes, see the Javadoc of defaultStatusHandler in WebClient.Builder.

The JmsTemplate class is the central class in the JMS core package. It simplifies the use of JMS, since it handles the creation and release of resources when sending or synchronously receiving messages.

Code that uses the JmsTemplate needs only to implement callback interfaces that give them a clearly defined high-level contract. The MessageCreator callback interface creates a message when given a Session provided by the calling code in JmsTemplate. To allow for more complex usage of the JMS API, SessionCallback provides the JMS session, and ProducerCallback exposes a Session and MessageProducer pair.

The JMS API exposes two types of send methods, one that takes delivery mode, priority, and time-to-live as Quality of Service (QOS) parameters and one that takes no QOS parameters and uses default values. Since JmsTemplate has many send methods, setting the QOS parameters have been exposed as bean properties to avoid duplication in the number of send methods. Similarly, the timeout value for synchronous receive calls is set by using the setReceiveTimeout property.

Some JMS providers allow the setting of default QOS values administratively through the configuration of the ConnectionFactory. This has the effect that a call to a MessageProducer instance’s send method (send(Destination destination, Message message)) uses different QOS default values than those specified in the JMS specification. In order to provide consistent management of QOS values, the JmsTemplate must, therefore, be specifically enabled to use its own QOS values by setting the boolean property isExplicitQosEnabled to true.

For convenience, JmsTemplate also exposes a basic request-reply operation that allows for sending a message and waiting for a reply on a temporary queue that is created as part of the operation.

As of Spring Framework 4.1, JmsMessagingTemplate is built on top of JmsTemplate and provides an integration with the messaging abstraction — that is, org.springframework.messaging.Message. This lets you create the message to send in a generic manner.

The JmsTemplate requires a reference to a ConnectionFactory. The ConnectionFactory is part of the JMS specification and serves as the entry point for working with JMS. It is used by the client application as a factory to create connections with the JMS provider and encapsulates various configuration parameters, many of which are vendor-specific, such as SSL configuration options.

When using JMS inside an EJB, the vendor provides implementations of the JMS interfaces so that they can participate in declarative transaction management and perform pooling of connections and sessions. In order to use this implementation, Jakarta EE containers typically require that you declare a JMS connection factory as a resource-ref inside the EJB or servlet deployment descriptors. To ensure the use of these features with the JmsTemplate inside an EJB, the client application should ensure that it references the managed implementation of the ConnectionFactory.

Destinations, as ConnectionFactory instances, are JMS administered objects that you can store and retrieve in JNDI. When configuring a Spring application context, you can use the JNDI JndiObjectFactoryBean factory class or <jee:jndi-lookup> to perform dependency injection on your object’s references to JMS destinations. However, this strategy is often cumbersome if there are a large number of destinations in the application or if there are advanced destination management features unique to the JMS provider. Examples of such advanced destination management include the creation of dynamic destinations or support for a hierarchical namespace of destinations. The JmsTemplate delegates the resolution of a destination name to a JMS destination object that implements the DestinationResolver interface. DynamicDestinationResolver is the default implementation used by JmsTemplate and accommodates resolving dynamic destinations. A JndiDestinationResolver is also provided to act as a service locator for destinations contained in JNDI and optionally falls back to the behavior contained in DynamicDestinationResolver.

Quite often, the destinations used in a JMS application are only known at runtime and, therefore, cannot be administratively created when the application is deployed. This is often because there is shared application logic between interacting system components that create destinations at runtime according to a well-known naming convention. Even though the creation of dynamic destinations is not part of the JMS specification, most vendors have provided this functionality. Dynamic destinations are created with a user-defined name, which differentiates them from temporary destinations, and are often not registered in JNDI. The API used to create dynamic destinations varies from provider to provider since the properties associated with the destination are vendor-specific. However, a simple implementation choice that is sometimes made by vendors is to disregard the warnings in the JMS specification and to use the method TopicSession createTopic(String topicName) or the QueueSession createQueue(String queueName) method to create a new destination with default destination properties. Depending on the vendor implementation, DynamicDestinationResolver can then also create a physical destination instead of only resolving one.

The boolean property pubSubDomain is used to configure the JmsTemplate with knowledge of what JMS domain is being used. By default, the value of this property is false, indicating that the point-to-point domain, Queues, is to be used. This property (used by JmsTemplate) determines the behavior of dynamic destination resolution through implementations of the DestinationResolver interface.

You can also configure the JmsTemplate with a default destination through the property defaultDestination. The default destination is with send and receive operations that do not refer to a specific destination.

One of the most common uses of JMS messages in the EJB world is to drive message-driven beans (MDBs). Spring offers a solution to create message-driven POJOs (MDPs) in a way that does not tie a user to an EJB container. (See Asynchronous reception: Message-Driven POJOs for detailed coverage of Spring’s MDP support.) Since Spring Framework 4.1, endpoint methods can be annotated with @JmsListener — see Annotation-driven Listener Endpoints for more details.

A message listener container is used to receive messages from a JMS message queue and drive the MessageListener that is injected into it. The listener container is responsible for all threading of message reception and dispatches into the listener for processing. A message listener container is the intermediary between an MDP and a messaging provider and takes care of registering to receive messages, participating in transactions, resource acquisition and release, exception conversion, and so on. This lets you write the (possibly complex) business logic associated with receiving a message (and possibly respond to it), and delegates boilerplate JMS infrastructure concerns to the framework.

There are two standard JMS message listener containers packaged with Spring, each with its specialized feature set.

Spring provides a JmsTransactionManager that manages transactions for a single JMS ConnectionFactory. This lets JMS applications leverage the managed-transaction features of Spring, as described in Transaction Management section of the Data Access chapter. The JmsTransactionManager performs local resource transactions, binding a JMS Connection/Session pair from the specified ConnectionFactory to the thread. JmsTemplate automatically detects such transactional resources and operates on them accordingly.

In a Jakarta EE environment, the ConnectionFactory pools Connection and Session instances, so those resources are efficiently reused across transactions. In a standalone environment, using Spring’s SingleConnectionFactory result in a shared JMS Connection, with each transaction having its own independent Session. Alternatively, consider the use of a provider-specific pooling adapter, such as ActiveMQ’s PooledConnectionFactory class.

You can also use JmsTemplate with the JtaTransactionManager and an XA-capable JMS ConnectionFactory to perform distributed transactions. Note that this requires the use of a JTA transaction manager as well as a properly XA-configured ConnectionFactory. (Check your Jakarta EE server’s or JMS provider’s documentation.)

Reusing code across a managed and unmanaged transactional environment can be confusing when using the JMS API to create a Session from a Connection. This is because the JMS API has only one factory method to create a Session, and it requires values for the transaction and acknowledgment modes. In a managed environment, setting these values is the responsibility of the environment’s transactional infrastructure, so these values are ignored by the vendor’s wrapper to the JMS Connection. When you use the JmsTemplate in an unmanaged environment, you can specify these values through the use of the properties sessionTransacted and sessionAcknowledgeMode. When you use a PlatformTransactionManager with JmsTemplate, the template is always given a transactional JMS Session.

To facilitate the sending of domain model objects, the JmsTemplate has various send methods that take a Java object as an argument for a message’s data content. The overloaded methods convertAndSend() and receiveAndConvert() methods in JmsTemplate delegate the conversion process to an instance of the MessageConverter interface. This interface defines a simple contract to convert between Java objects and JMS messages. The default implementation (SimpleMessageConverter) supports conversion between String and TextMessage, byte[] and BytesMessage, and java.util.Map and MapMessage. By using the converter, you and your application code can focus on the business object that is being sent or received through JMS and not be concerned with the details of how it is represented as a JMS message.

The sandbox currently includes a MapMessageConverter, which uses reflection to convert between a JavaBean and a MapMessage. Other popular implementation choices you might implement yourself are converters that use an existing XML marshalling package (such as JAXB or XStream) to create a TextMessage that represents the object.

To accommodate the setting of a message’s properties, headers, and body that can not be generically encapsulated inside a converter class, the MessagePostProcessor interface gives you access to the message after it has been converted but before it is sent. The following example shows how to modify a message header and a property after a java.util.Map is converted to a message:

This results in a message of the following form:

While the send operations cover many common usage scenarios, you might sometimes want to perform multiple operations on a JMS Session or MessageProducer. The SessionCallback and ProducerCallback expose the JMS Session and Session / MessageProducer pair, respectively. The execute() methods on JmsTemplate run these callback methods.

While JMS is typically associated with asynchronous processing, you can consume messages synchronously. The overloaded receive(..) methods provide this functionality. During a synchronous receive, the calling thread blocks until a message becomes available. This can be a dangerous operation, since the calling thread can potentially be blocked indefinitely. The receiveTimeout property specifies how long the receiver should wait before giving up waiting for a message.

In a fashion similar to a Message-Driven Bean (MDB) in the EJB world, the Message-Driven POJO (MDP) acts as a receiver for JMS messages. The one restriction (but see Using MessageListenerAdapter) on an MDP is that it must implement the jakarta.jms.MessageListener interface. Note that, if your POJO receives messages on multiple threads, it is important to ensure that your implementation is thread-safe.

The following example shows a simple implementation of an MDP:

Once you have implemented your MessageListener, it is time to create a message listener container.

The following example shows how to define and configure one of the message listener containers that ships with Spring (in this case, DefaultMessageListenerContainer):

See the Spring javadoc of the various message listener containers (all of which implement MessageListenerContainer) for a full description of the features supported by each implementation.

The SessionAwareMessageListener interface is a Spring-specific interface that provides a similar contract to the JMS MessageListener interface but also gives the message-handling method access to the JMS Session from which the Message was received. The following listing shows the definition of the SessionAwareMessageListener interface:

You can choose to have your MDPs implement this interface (in preference to the standard JMS MessageListener interface) if you want your MDPs to be able to respond to any received messages (by using the Session supplied in the onMessage(Message, Session) method). All of the message listener container implementations that ship with Spring have support for MDPs that implement either the MessageListener or SessionAwareMessageListener interface. Classes that implement the SessionAwareMessageListener come with the caveat that they are then tied to Spring through the interface. The choice of whether or not to use it is left entirely up to you as an application developer or architect.

Note that the onMessage(..) method of the SessionAwareMessageListener interface throws JMSException. In contrast to the standard JMS MessageListener interface, when using the SessionAwareMessageListener interface, it is the responsibility of the client code to handle any thrown exceptions.

The MessageListenerAdapter class is the final component in Spring’s asynchronous messaging support. In a nutshell, it lets you expose almost any class as an MDP (though there are some constraints).

Consider the following interface definition:

Notice that, although the interface extends neither the MessageListener nor the SessionAwareMessageListener interface, you can still use it as an MDP by using the MessageListenerAdapter class. Notice also how the various message handling methods are strongly typed according to the contents of the various Message types that they can receive and handle.

Now consider the following implementation of the MessageDelegate interface:

In particular, note how the preceding implementation of the MessageDelegate interface (the DefaultMessageDelegate class) has no JMS dependencies at all. It truly is a POJO that we can make into an MDP through the following configuration:

The next example shows another MDP that can handle only receiving JMS TextMessage messages. Notice how the message handling method is actually called receive (the name of the message handling method in a MessageListenerAdapter defaults to handleMessage), but it is configurable (as you can see later in this section). Notice also how the receive(..) method is strongly typed to receive and respond only to JMS TextMessage messages. The following listing shows the definition of the TextMessageDelegate interface:

The following listing shows a class that implements the TextMessageDelegate interface:

The configuration of the attendant MessageListenerAdapter would then be as follows:

Note that, if the messageListener receives a JMS Message of a type other than TextMessage, an IllegalStateException is thrown (and subsequently swallowed). Another of the capabilities of the MessageListenerAdapter class is the ability to automatically send back a response Message if a handler method returns a non-void value. Consider the following interface and class:

If you use the DefaultResponsiveTextMessageDelegate in conjunction with a MessageListenerAdapter, any non-null value that is returned from the execution of the 'receive(..)' method is (in the default configuration) converted into a TextMessage. The resulting TextMessage is then sent to the Destination (if one exists) defined in the JMS Reply-To property of the original Message or the default Destination set on the MessageListenerAdapter (if one has been configured). If no Destination is found, an InvalidDestinationException is thrown (note that this exception is not swallowed and propagates up the call stack).

Invoking a message listener within a transaction requires only reconfiguration of the listener container.

You can activate local resource transactions through the sessionTransacted flag on the listener container definition. Each message listener invocation then operates within an active JMS transaction, with message reception rolled back in case of listener execution failure. Sending a response message (through SessionAwareMessageListener) is part of the same local transaction, but any other resource operations (such as database access) operate independently. This usually requires duplicate message detection in the listener implementation, to cover the case where database processing has committed but message processing failed to commit.

Consider the following bean definition:

To participate in an externally managed transaction, you need to configure a transaction manager and use a listener container that supports externally managed transactions (typically, DefaultMessageListenerContainer).

To configure a message listener container for XA transaction participation, you want to configure a JtaTransactionManager (which, by default, delegates to the Jakarta EE server’s transaction subsystem). Note that the underlying JMS ConnectionFactory needs to be XA-capable and properly registered with your JTA transaction coordinator. (Check your Jakarta EE server’s configuration of JNDI resources.) This lets message reception as well as (for example) database access be part of the same transaction (with unified commit semantics, at the expense of XA transaction log overhead).

The following bean definition creates a transaction manager:

Then we need to add it to our earlier container configuration. The container takes care of the rest. The following example shows how to do so:

To enable support for @JmsListener annotations, you can add @EnableJms to one of your @Configuration classes, as the following example shows:

By default, the infrastructure looks for a bean named jmsListenerContainerFactory as the source for the factory to use to create message listener containers. In this case (and ignoring the JMS infrastructure setup), you can invoke the processOrder method with a core poll size of three threads and a maximum pool size of ten threads.

You can customize the listener container factory to use for each annotation or you can configure an explicit default by implementing the JmsListenerConfigurer interface. The default is required only if at least one endpoint is registered without a specific container factory. See the javadoc of classes that implement JmsListenerConfigurer for details and examples.

If you prefer XML configuration, you can use the <jms:annotation-driven> element, as the following example shows:

JmsListenerEndpoint provides a model of a JMS endpoint and is responsible for configuring the container for that model. The infrastructure lets you programmatically configure endpoints in addition to the ones that are detected by the JmsListener annotation. The following example shows how to do so:

In the preceding example, we used SimpleJmsListenerEndpoint, which provides the actual MessageListener to invoke. However, you could also build your own endpoint variant to describe a custom invocation mechanism.

Note that you could skip the use of @JmsListener altogether and programmatically register only your endpoints through JmsListenerConfigurer.

So far, we have been injecting a simple String in our endpoint, but it can actually have a very flexible method signature. In the following example, we rewrite it to inject the Order with a custom header:

The main elements you can inject in JMS listener endpoints are as follows:

The ability to inject Spring’s Message abstraction is particularly useful to benefit from all the information stored in the transport-specific message without relying on transport-specific API. The following example shows how to do so:

Handling of method arguments is provided by DefaultMessageHandlerMethodFactory, which you can further customize to support additional method arguments. You can customize the conversion and validation support there as well.

For instance, if we want to make sure our Order is valid before processing it, we can annotate the payload with @Valid and configure the necessary validator, as the following example shows:

The existing support in MessageListenerAdapter already lets your method have a non-void return type. When that is the case, the result of the invocation is encapsulated in a jakarta.jms.Message, sent either in the destination specified in the JMSReplyTo header of the original message or in the default destination configured on the listener. You can now set that default destination by using the @SendTo annotation of the messaging abstraction.

Assuming that our processOrder method should now return an OrderStatus, we can write it to automatically send a response, as the following example shows:

If you need to set additional headers in a transport-independent manner, you can return a Message instead, with a method similar to the following:

If you need to compute the response destination at runtime, you can encapsulate your response in a JmsResponse instance that also provides the destination to use at runtime. We can rewrite the previous example as follows:

Finally, if you need to specify some QoS values for the response such as the priority or the time to live, you can configure the JmsListenerContainerFactory accordingly, as the following example shows:

The configuration shown in the preceding section assumes that the application is running in an environment that has one (and only one) MBeanServer already running. In this case, Spring tries to locate the running MBeanServer and register your beans with that server (if any). This behavior is useful when your application runs inside a container (such as Tomcat or IBM WebSphere) that has its own MBeanServer.

However, this approach is of no use in a standalone environment or when running inside a container that does not provide an MBeanServer. To address this, you can create an MBeanServer instance declaratively by adding an instance of the org.springframework.jmx.support.MBeanServerFactoryBean class to your configuration. You can also ensure that a specific MBeanServer is used by setting the value of the MBeanExporter instance’s server property to the MBeanServer value returned by an MBeanServerFactoryBean, as the following example shows:

In the preceding example, an instance of MBeanServer is created by the MBeanServerFactoryBean and is supplied to the MBeanExporter through the server property. When you supply your own MBeanServer instance, the MBeanExporter does not try to locate a running MBeanServer and uses the supplied MBeanServer instance. For this to work correctly, you must have a JMX implementation on your classpath.

If no server is specified, the MBeanExporter tries to automatically detect a running MBeanServer. This works in most environments, where only one MBeanServer instance is used. However, when multiple instances exist, the exporter might pick the wrong server. In such cases, you should use the MBeanServer agentId to indicate which instance to be used, as the following example shows:

For platforms or cases where the existing MBeanServer has a dynamic (or unknown) agentId that is retrieved through lookup methods, you should use factory-method, as the following example shows:

If you configure a bean with an MBeanExporter that is also configured for lazy initialization, the MBeanExporter does not break this contract and avoids instantiating the bean. Instead, it registers a proxy with the MBeanServer and defers obtaining the bean from the container until the first invocation on the proxy occurs.

Any beans that are exported through the MBeanExporter and are already valid MBeans are registered as-is with the MBeanServer without further intervention from Spring. You can cause MBeans to be automatically detected by the MBeanExporter by setting the autodetect property to true, as the following example shows:

In the preceding example, the bean called spring:mbean=true is already a valid JMX MBean and is automatically registered by Spring. By default, a bean that is autodetected for JMX registration has its bean name used as the ObjectName. You can override this behavior, as detailed in Controlling ObjectName Instances for Your Beans.

Consider the scenario where a Spring MBeanExporter attempts to register an MBean with an MBeanServer by using the ObjectName bean:name=testBean1. If an MBean instance has already been registered under that same ObjectName, the default behavior is to fail (and throw an InstanceAlreadyExistsException).

You can control exactly what happens when an MBean is registered with an MBeanServer. Spring’s JMX support allows for three different registration behaviors to control the registration behavior when the registration process finds that an MBean has already been registered under the same ObjectName. The following table summarizes these registration behaviors:

The values in the preceding table are defined as enums on the RegistrationPolicy class. If you want to change the default registration behavior, you need to set the value of the registrationPolicy property on your MBeanExporter definition to one of those values.

The following example shows how to change from the default registration behavior to the REPLACE_EXISTING behavior:

Behind the scenes, the MBeanExporter delegates to an implementation of the org.springframework.jmx.export.assembler.MBeanInfoAssembler interface, which is responsible for defining the management interface of each bean that is exposed. The default implementation, org.springframework.jmx.export.assembler.SimpleReflectiveMBeanInfoAssembler, defines a management interface that exposes all public properties and methods (as you saw in the examples in the preceding sections). Spring provides two additional implementations of the MBeanInfoAssembler interface that let you control the generated management interface by using either source-level metadata or any arbitrary interface.

By using the MetadataMBeanInfoAssembler, you can define the management interfaces for your beans by using source-level metadata. The reading of metadata is encapsulated by the org.springframework.jmx.export.metadata.JmxAttributeSource interface. Spring JMX provides a default implementation that uses Java annotations, namely org.springframework.jmx.export.annotation.AnnotationJmxAttributeSource. You must configure the MetadataMBeanInfoAssembler with an implementation instance of the JmxAttributeSource interface for it to function correctly (there is no default).

To mark a bean for export to JMX, you should annotate the bean class with the ManagedResource annotation. You must mark each method you wish to expose as an operation with the ManagedOperation annotation and mark each property you wish to expose with the ManagedAttribute annotation. When marking properties, you can omit either the annotation of the getter or the setter to create a write-only or read-only attribute, respectively.

The following example shows the annotated version of the JmxTestBean class that we used in Creating an MBeanServer:

In the preceding example, you can see that the JmxTestBean class is marked with the ManagedResource annotation and that this ManagedResource annotation is configured with a set of properties. These properties can be used to configure various aspects of the MBean that is generated by the MBeanExporter and are explained in greater detail later in Source-level Metadata Types.

Both the age and name properties are annotated with the ManagedAttribute annotation, but, in the case of the age property, only the getter is marked. This causes both of these properties to be included in the management interface as attributes, but the age attribute is read-only.

Finally, the add(int, int) method is marked with the ManagedOperation attribute, whereas the dontExposeMe() method is not. This causes the management interface to contain only one operation (add(int, int)) when you use the MetadataMBeanInfoAssembler.

The following configuration shows how you can configure the MBeanExporter to use the MetadataMBeanInfoAssembler:

In the preceding example, an MetadataMBeanInfoAssembler bean has been configured with an instance of the AnnotationJmxAttributeSource class and passed to the MBeanExporter through the assembler property. This is all that is required to take advantage of metadata-driven management interfaces for your Spring-exposed MBeans.

The following table describes the source-level metadata types that are available for use in Spring JMX:

The following table describes the configuration parameters that are available for use on these source-level metadata types:

To simplify configuration even further, Spring includes the AutodetectCapableMBeanInfoAssembler interface, which extends the MBeanInfoAssembler interface to add support for autodetection of MBean resources. If you configure the MBeanExporter with an instance of AutodetectCapableMBeanInfoAssembler, it is allowed to “vote” on the inclusion of beans for exposure to JMX.

The only implementation of the AutodetectCapableMBeanInfo interface is the MetadataMBeanInfoAssembler, which votes to include any bean that is marked with the ManagedResource attribute. The default approach in this case is to use the bean name as the ObjectName, which results in a configuration similar to the following:

Notice that, in the preceding configuration, no beans are passed to the MBeanExporter. However, the JmxTestBean is still registered, since it is marked with the ManagedResource attribute and the MetadataMBeanInfoAssembler detects this and votes to include it. The only problem with this approach is that the name of the JmxTestBean now has business meaning. You can address this issue by changing the default behavior for ObjectName creation as defined in Controlling ObjectName Instances for Your Beans.

In addition to the MetadataMBeanInfoAssembler, Spring also includes the InterfaceBasedMBeanInfoAssembler, which lets you constrain the methods and properties that are exposed based on the set of methods defined in a collection of interfaces.

Although the standard mechanism for exposing MBeans is to use interfaces and a simple naming scheme, InterfaceBasedMBeanInfoAssembler extends this functionality by removing the need for naming conventions, letting you use more than one interface and removing the need for your beans to implement the MBean interfaces.

Consider the following interface, which is used to define a management interface for the JmxTestBean class that we showed earlier:

This interface defines the methods and properties that are exposed as operations and attributes on the JMX MBean. The following code shows how to configure Spring JMX to use this interface as the definition for the management interface:

In the preceding example, the InterfaceBasedMBeanInfoAssembler is configured to use the IJmxTestBean interface when constructing the management interface for any bean. It is important to understand that beans processed by the InterfaceBasedMBeanInfoAssembler are not required to implement the interface used to generate the JMX management interface.

In the preceding case, the IJmxTestBean interface is used to construct all management interfaces for all beans. In many cases, this is not the desired behavior, and you may want to use different interfaces for different beans. In this case, you can pass InterfaceBasedMBeanInfoAssembler a Properties instance through the interfaceMappings property, where the key of each entry is the bean name and the value of each entry is a comma-separated list of interface names to use for that bean.

If no management interface is specified through either the managedInterfaces or interfaceMappings properties, the InterfaceBasedMBeanInfoAssembler reflects on the bean and uses all of the interfaces implemented by that bean to create the management interface.

MethodNameBasedMBeanInfoAssembler lets you specify a list of method names that are exposed to JMX as attributes and operations. The following code shows a sample configuration:

In the preceding example, you can see that the add and myOperation methods are exposed as JMX operations, and getName(), setName(String), and getAge() are exposed as the appropriate half of a JMX attribute. In the preceding code, the method mappings apply to beans that are exposed to JMX. To control method exposure on a bean-by-bean basis, you can use the methodMappings property of MethodNameMBeanInfoAssembler to map bean names to lists of method names.

You can configure your own KeyNamingStrategy instance and configure it to read ObjectName instances from a Properties instance rather than use a bean key. The KeyNamingStrategy tries to locate an entry in the Properties with a key that corresponds to the bean key. If no entry is found or if the Properties instance is null, the bean key itself is used.

The following code shows a sample configuration for the KeyNamingStrategy:

The preceding example configures an instance of KeyNamingStrategy with a Properties instance that is merged from the Properties instance defined by the mapping property and the properties files located in the paths defined by the mappings property. In this configuration, the testBean bean is given an ObjectName of bean:name=testBean1, since this is the entry in the Properties instance that has a key corresponding to the bean key.

If no entry in the Properties instance can be found, the bean key name is used as the ObjectName.

MetadataNamingStrategy uses the objectName property of the ManagedResource attribute on each bean to create the ObjectName. The following code shows the configuration for the MetadataNamingStrategy:

If no objectName has been provided for the ManagedResource attribute, an ObjectName is created with the following format: [fully-qualified-package-name]:type=[short-classname],name=[bean-name]. For example, the generated ObjectName for the following bean would be com.example:type=MyClass,name=myBean:

If you prefer to use the annotation-based approach to define your management interfaces, a convenience subclass of MBeanExporter is available: AnnotationMBeanExporter. When defining an instance of this subclass, you no longer need the namingStrategy, assembler, and attributeSource configuration, since it always uses standard Java annotation-based metadata (autodetection is always enabled as well). In fact, rather than defining an MBeanExporter bean, an even simpler syntax is supported by the @EnableMBeanExport @Configuration annotation, as the following example shows:

If you prefer XML-based configuration, the <context:mbean-export/> element serves the same purpose and is shown in the following listing:

If necessary, you can provide a reference to a particular MBean server, and the defaultDomain attribute (a property of AnnotationMBeanExporter) accepts an alternate value for the generated MBean ObjectName domains. This is used in place of the fully qualified package name as described in the previous section on MetadataNamingStrategy, as the following example shows:

The following example shows the XML equivalent of the preceding annotation-based example:

To have Spring JMX create, start, and expose a JSR-160 JMXConnectorServer, you can use the following configuration:

By default, ConnectorServerFactoryBean creates a JMXConnectorServer bound to service:jmx:jmxmp://localhost:9875. The serverConnector bean thus exposes the local MBeanServer to clients through the JMXMP protocol on localhost, port 9875. Note that the JMXMP protocol is marked as optional by the JSR 160 specification. Currently, the main open-source JMX implementation, MX4J, and the one provided with the JDK do not support JMXMP.

To specify another URL and register the JMXConnectorServer itself with the MBeanServer, you can use the serviceUrl and ObjectName properties, respectively, as the following example shows:

If the ObjectName property is set, Spring automatically registers your connector with the MBeanServer under that ObjectName. The following example shows the full set of parameters that you can pass to the ConnectorServerFactoryBean when creating a JMXConnector:

Note that, when you use a RMI-based connector, you need the lookup service (tnameserv or rmiregistry) to be started in order for the name registration to complete.

To create an MBeanServerConnection to a remote JSR-160-enabled MBeanServer, you can use the MBeanServerConnectionFactoryBean, as the following example shows:

JSR-160 permits extensions to the way in which communication is done between the client and the server. The examples shown in the preceding sections use the mandatory RMI-based implementation required by the JSR-160 specification (IIOP and JRMP) and the (optional) JMXMP. By using other providers or JMX implementations (such as MX4J) you can take advantage of protocols such as SOAP or Hessian over simple HTTP or SSL and others, as the following example shows:

In the preceding example, we used MX4J 3.0.0. See the official MX4J documentation for more information.

Spring’s JMX support makes it easy to register any number of NotificationListeners with any number of MBeans (this includes MBeans exported by Spring’s MBeanExporter and MBeans registered through some other mechanism). For example, consider the scenario where one would like to be informed (through a Notification) each and every time an attribute of a target MBean changes. The following example writes notifications to the console:

The following example adds ConsoleLoggingNotificationListener (defined in the preceding example) to notificationListenerMappings:

With the preceding configuration in place, every time a JMX Notification is broadcast from the target MBean (bean:name=testBean1), the ConsoleLoggingNotificationListener bean that was registered as a listener through the notificationListenerMappings property is notified. The ConsoleLoggingNotificationListener bean can then take whatever action it deems appropriate in response to the Notification.

You can also use straight bean names as the link between exported beans and listeners, as the following example shows:

If you want to register a single NotificationListener instance for all of the beans that the enclosing MBeanExporter exports, you can use the special wildcard (*) as the key for an entry in the notificationListenerMappings property map, as the following example shows:

If you need to do the inverse (that is, register a number of distinct listeners against an MBean), you must instead use the notificationListeners list property (in preference to the notificationListenerMappings property). This time, instead of configuring a NotificationListener for a single MBean, we configure NotificationListenerBean instances. A NotificationListenerBean encapsulates a NotificationListener and the ObjectName (or ObjectNames) that it is to be registered against in an MBeanServer. The NotificationListenerBean also encapsulates a number of other properties, such as a NotificationFilter and an arbitrary handback object that can be used in advanced JMX notification scenarios.

The configuration when using NotificationListenerBean instances is not wildly different to what was presented previously, as the following example shows:

The preceding example is equivalent to the first notification example. Assume, then, that we want to be given a handback object every time a Notification is raised and that we also want to filter out extraneous Notifications by supplying a NotificationFilter. The following example accomplishes these goals:

(For a full discussion of what a handback object is and, indeed, what a NotificationFilter is, see the section of the JMX specification (1.2) entitled 'The JMX Notification Model'.)

Spring provides support not only for registering to receive Notifications but also for publishing Notifications.

The key interface in Spring’s JMX notification publication support is the NotificationPublisher interface (defined in the org.springframework.jmx.export.notification package). Any bean that is going to be exported as an MBean through an MBeanExporter instance can implement the related NotificationPublisherAware interface to gain access to a NotificationPublisher instance. The NotificationPublisherAware interface supplies an instance of a NotificationPublisher to the implementing bean through a simple setter method, which the bean can then use to publish Notifications.

As stated in the javadoc of the NotificationPublisher interface, managed beans that publish events through the NotificationPublisher mechanism are not responsible for the state management of notification listeners. Spring’s JMX support takes care of handling all the JMX infrastructure issues. All you need to do, as an application developer, is implement the NotificationPublisherAware interface and start publishing events by using the supplied NotificationPublisher instance. Note that the NotificationPublisher is set after the managed bean has been registered with an MBeanServer.

Using a NotificationPublisher instance is quite straightforward. You create a JMX Notification instance (or an instance of an appropriate Notification subclass), populate the notification with the data pertinent to the event that is to be published, and invoke the sendNotification(Notification) on the NotificationPublisher instance, passing in the Notification.

In the following example, exported instances of the JmxTestBean publish a NotificationEvent every time the add(int, int) operation is invoked:

The NotificationPublisher interface and the machinery to get it all working is one of the nicer features of Spring’s JMX support. It does, however, come with the price tag of coupling your classes to both Spring and JMX. As always, the advice here is to be pragmatic. If you need the functionality offered by the NotificationPublisher and you can accept the coupling to both Spring and JMX, then do so.

The following example shows how to use MailSender and SimpleMailMessage to send an email when someone places an order:

The following example shows the bean definitions for the preceding code:

This section describes another implementation of OrderManager that uses the MimeMessagePreparator callback interface. In the following example, the mailSender property is of type JavaMailSender so that we are able to use the JavaMail MimeMessage class:

The Spring Framework’s mail support ships with the standard JavaMail implementation. See the relevant javadoc for more information.

Multipart email messages allow for both attachments and inline resources. Examples of inline resources include an image or a stylesheet that you want to use in your message but that you do not want displayed as an attachment.

The code in the examples shown in the previous sections explicitly created the content of the email message, by using methods calls such as message.setText(..). This is fine for simple cases, and it is okay in the context of the aforementioned examples, where the intent was to show you the very basics of the API.

In your typical enterprise application, though, developers often do not create the content of email messages by using the previously shown approach for a number of reasons:

Typically, the approach taken to address these issues is to use a template library (such as FreeMarker) to define the display structure of email content. This leaves your code tasked only with creating the data that is to be rendered in the email template and sending the email. It is definitely a best practice when the content of your email messages becomes even moderately complex, and, with the Spring Framework’s support classes for FreeMarker, it becomes quite easy to do.

Spring includes a number of pre-built implementations of TaskExecutor. In all likelihood, you should never need to implement your own. The variants that Spring provides are as follows:

Spring’s TaskExecutor implementations are used as simple JavaBeans. In the following example, we define a bean that uses the ThreadPoolTaskExecutor to asynchronously print out a set of messages:

As you can see, rather than retrieving a thread from the pool and executing it yourself, you add your Runnable to the queue. Then the TaskExecutor uses its internal rules to decide when the task gets run.

To configure the rules that the TaskExecutor uses, we expose simple bean properties:

The Trigger interface is essentially inspired by JSR-236 which, as of Spring 3.0, was not yet officially implemented. The basic idea of the Trigger is that execution times may be determined based on past execution outcomes or even arbitrary conditions. If these determinations do take into account the outcome of the preceding execution, that information is available within a TriggerContext. The Trigger interface itself is quite simple, as the following listing shows:

The TriggerContext is the most important part. It encapsulates all of the relevant data and is open for extension in the future, if necessary. The TriggerContext is an interface (a SimpleTriggerContext implementation is used by default). The following listing shows the available methods for Trigger implementations.

Spring provides two implementations of the Trigger interface. The most interesting one is the CronTrigger. It enables the scheduling of tasks based on cron expressions. For example, the following task is scheduled to run 15 minutes past each hour but only during the 9-to-5 “business hours” on weekdays:

The other implementation is a PeriodicTrigger that accepts a fixed period, an optional initial delay value, and a boolean to indicate whether the period should be interpreted as a fixed-rate or a fixed-delay. Since the TaskScheduler interface already defines methods for scheduling tasks at a fixed rate or with a fixed delay, those methods should be used directly whenever possible. The value of the PeriodicTrigger implementation is that you can use it within components that rely on the Trigger abstraction. For example, it may be convenient to allow periodic triggers, cron-based triggers, and even custom trigger implementations to be used interchangeably. Such a component could take advantage of dependency injection so that you can configure such Triggers externally and, therefore, easily modify or extend them.

As with Spring’s TaskExecutor abstraction, the primary benefit of the TaskScheduler arrangement is that an application’s scheduling needs are decoupled from the deployment environment. This abstraction level is particularly relevant when deploying to an application server environment where threads should not be created directly by the application itself. For such scenarios, Spring provides a TimerManagerTaskScheduler that delegates to a CommonJ TimerManager on WebLogic or WebSphere as well as a more recent DefaultManagedTaskScheduler that delegates to a JSR-236 ManagedScheduledExecutorService in a Jakarta EE environment. Both are typically configured with a JNDI lookup.

Whenever external thread management is not a requirement, a simpler alternative is a local ScheduledExecutorService setup within the application, which can be adapted through Spring’s ConcurrentTaskScheduler. As a convenience, Spring also provides a ThreadPoolTaskScheduler, which internally delegates to a ScheduledExecutorService to provide common bean-style configuration along the lines of ThreadPoolTaskExecutor. These variants work perfectly fine for locally embedded thread pool setups in lenient application server environments, as well — in particular on Tomcat and Jetty.

To enable support for @Scheduled and @Async annotations, you can add @EnableScheduling and @EnableAsync to one of your @Configuration classes, as the following example shows:

You can pick and choose the relevant annotations for your application. For example, if you need only support for @Scheduled, you can omit @EnableAsync. For more fine-grained control, you can additionally implement the SchedulingConfigurer interface, the AsyncConfigurer interface, or both. See the SchedulingConfigurer and AsyncConfigurer javadoc for full details.

If you prefer XML configuration, you can use the <task:annotation-driven> element, as the following example shows:

Note that, with the preceding XML, an executor reference is provided for handling those tasks that correspond to methods with the @Async annotation, and the scheduler reference is provided for managing those methods annotated with @Scheduled.

You can add the @Scheduled annotation to a method, along with trigger metadata. For example, the following method is invoked every five seconds (5000 milliseconds) with a fixed delay, meaning that the period is measured from the completion time of each preceding invocation.

If you need a fixed-rate execution, you can use the fixedRate attribute within the annotation. The following method is invoked every five seconds (measured between the successive start times of each invocation).

For fixed-delay and fixed-rate tasks, you can specify an initial delay by indicating the amount of time to wait before the first execution of the method, as the following fixedRate example shows.

If simple periodic scheduling is not expressive enough, you can provide a cron expression. The following example runs only on weekdays:

Notice that the methods to be scheduled must have void returns and must not accept any arguments. If the method needs to interact with other objects from the application context, those would typically have been provided through dependency injection.

You can provide the @Async annotation on a method so that invocation of that method occurs asynchronously. In other words, the caller returns immediately upon invocation, while the actual execution of the method occurs in a task that has been submitted to a Spring TaskExecutor. In the simplest case, you can apply the annotation to a method that returns void, as the following example shows:

Unlike the methods annotated with the @Scheduled annotation, these methods can expect arguments, because they are invoked in the “normal” way by callers at runtime rather than from a scheduled task being managed by the container. For example, the following code is a legitimate application of the @Async annotation:

Even methods that return a value can be invoked asynchronously. However, such methods are required to have a Future-typed return value. This still provides the benefit of asynchronous execution so that the caller can perform other tasks prior to calling get() on that Future. The following example shows how to use @Async on a method that returns a value:

You can not use @Async in conjunction with lifecycle callbacks such as @PostConstruct. To asynchronously initialize Spring beans, you currently have to use a separate initializing Spring bean that then invokes the @Async annotated method on the target, as the following example shows:

By default, when specifying @Async on a method, the executor that is used is the one configured when enabling async support, i.e. the “annotation-driven” element if you are using XML or your AsyncConfigurer implementation, if any. However, you can use the value attribute of the @Async annotation when you need to indicate that an executor other than the default should be used when executing a given method. The following example shows how to do so:

In this case, "otherExecutor" can be the name of any Executor bean in the Spring container, or it may be the name of a qualifier associated with any Executor (for example, as specified with the <qualifier> element or Spring’s @Qualifier annotation).

When an @Async method has a Future-typed return value, it is easy to manage an exception that was thrown during the method execution, as this exception is thrown when calling get on the Future result. With a void return type, however, the exception is uncaught and cannot be transmitted. You can provide an AsyncUncaughtExceptionHandler to handle such exceptions. The following example shows how to do so:

By default, the exception is merely logged. You can define a custom AsyncUncaughtExceptionHandler by using AsyncConfigurer or the <task:annotation-driven/> XML element.

The following element creates a ThreadPoolTaskScheduler instance with the specified thread pool size:

The value provided for the id attribute is used as the prefix for thread names within the pool. The scheduler element is relatively straightforward. If you do not provide a pool-size attribute, the default thread pool has only a single thread. There are no other configuration options for the scheduler.

The following creates a ThreadPoolTaskExecutor instance:

As with the scheduler shown in the previous section, the value provided for the id attribute is used as the prefix for thread names within the pool. As far as the pool size is concerned, the executor element supports more configuration options than the scheduler element. For one thing, the thread pool for a ThreadPoolTaskExecutor is itself more configurable. Rather than only a single size, an executor’s thread pool can have different values for the core and the max size. If you provide a single value, the executor has a fixed-size thread pool (the core and max sizes are the same). However, the executor element’s pool-size attribute also accepts a range in the form of min-max. The following example sets a minimum value of 5 and a maximum value of 25:

In the preceding configuration, a queue-capacity value has also been provided. The configuration of the thread pool should also be considered in light of the executor’s queue capacity. For the full description of the relationship between pool size and queue capacity, see the documentation for ThreadPoolExecutor. The main idea is that, when a task is submitted, the executor first tries to use a free thread if the number of active threads is currently less than the core size. If the core size has been reached, the task is added to the queue, as long as its capacity has not yet been reached. Only then, if the queue’s capacity has been reached, does the executor create a new thread beyond the core size. If the max size has also been reached, then the executor rejects the task.

By default, the queue is unbounded, but this is rarely the desired configuration, because it can lead to OutOfMemoryErrors if enough tasks are added to that queue while all pool threads are busy. Furthermore, if the queue is unbounded, the max size has no effect at all. Since the executor always tries the queue before creating a new thread beyond the core size, a queue must have a finite capacity for the thread pool to grow beyond the core size (this is why a fixed-size pool is the only sensible case when using an unbounded queue).

Consider the case, as mentioned above, when a task is rejected. By default, when a task is rejected, a thread pool executor throws a TaskRejectedException. However, the rejection policy is actually configurable. The exception is thrown when using the default rejection policy, which is the AbortPolicy implementation. For applications where some tasks can be skipped under heavy load, you can instead configure either DiscardPolicy or DiscardOldestPolicy. Another option that works well for applications that need to throttle the submitted tasks under heavy load is the CallerRunsPolicy. Instead of throwing an exception or discarding tasks, that policy forces the thread that is calling the submit method to run the task itself. The idea is that such a caller is busy while running that task and not able to submit other tasks immediately. Therefore, it provides a simple way to throttle the incoming load while maintaining the limits of the thread pool and queue. Typically, this allows the executor to “catch up” on the tasks it is handling and thereby frees up some capacity on the queue, in the pool, or both. You can choose any of these options from an enumeration of values available for the rejection-policy attribute on the executor element.

The following example shows an executor element with a number of attributes to specify various behaviors:

Finally, the keep-alive setting determines the time limit (in seconds) for which threads may remain idle before being stopped. If there are more than the core number of threads currently in the pool, after waiting this amount of time without processing a task, excess threads get stopped. A time value of zero causes excess threads to stop immediately after executing a task without remaining follow-up work in the task queue. The following example sets the keep-alive value to two minutes:

The most powerful feature of Spring’s task namespace is the support for configuring tasks to be scheduled within a Spring Application Context. This follows an approach similar to other “method-invokers” in Spring, such as that provided by the JMS namespace for configuring message-driven POJOs. Basically, a ref attribute can point to any Spring-managed object, and the method attribute provides the name of a method to be invoked on that object. The following listing shows a simple example:

The scheduler is referenced by the outer element, and each individual task includes the configuration of its trigger metadata. In the preceding example, that metadata defines a periodic trigger with a fixed delay indicating the number of milliseconds to wait after each task execution has completed. Another option is fixed-rate, indicating how often the method should be run regardless of how long any previous execution takes. Additionally, for both fixed-delay and fixed-rate tasks, you can specify an 'initial-delay' parameter, indicating the number of milliseconds to wait before the first execution of the method. For more control, you can instead provide a cron attribute to provide a cron expression. The following example shows these other options:

Expressions such as 0 0 * * * * are hard for humans to parse and are, therefore, hard to fix in case of bugs. To improve readability, Spring supports the following macros, which represent commonly used sequences. You can use these macros instead of the six-digit value, thus: @Scheduled(cron = "@hourly").

Quartz JobDetail objects contain all the information needed to run a job. Spring provides a JobDetailFactoryBean, which provides bean-style properties for XML configuration purposes. Consider the following example:

The job detail configuration has all the information it needs to run the job (ExampleJob). The timeout is specified in the job data map. The job data map is available through the JobExecutionContext (passed to you at execution time), but the JobDetail also gets its properties from the job data mapped to properties of the job instance. So, in the following example, the ExampleJob contains a bean property named timeout, and the JobDetail has it applied automatically:

All additional properties from the job data map are available to you as well.

Often you merely need to invoke a method on a specific object. By using the MethodInvokingJobDetailFactoryBean, you can do exactly this, as the following example shows:

The preceding example results in the doIt method being called on the exampleBusinessObject method, as the following example shows:

By using the MethodInvokingJobDetailFactoryBean, you need not create one-line jobs that merely invoke a method. You need only create the actual business object and wire up the detail object.

By default, Quartz Jobs are stateless, resulting in the possibility of jobs interfering with each other. If you specify two triggers for the same JobDetail, it is possible that, before the first job has finished, the second one starts. If JobDetail classes implement the Stateful interface, this does not happen. The second job does not start before the first one has finished. To make jobs resulting from the MethodInvokingJobDetailFactoryBean be non-concurrent, set the concurrent flag to false, as the following example shows:

We have created job details and jobs. We have also reviewed the convenience bean that lets you invoke a method on a specific object. Of course, we still need to schedule the jobs themselves. This is done by using triggers and a SchedulerFactoryBean. Several triggers are available within Quartz, and Spring offers two Quartz FactoryBean implementations with convenient defaults: CronTriggerFactoryBean and SimpleTriggerFactoryBean.

Triggers need to be scheduled. Spring offers a SchedulerFactoryBean that exposes triggers to be set as properties. SchedulerFactoryBean schedules the actual jobs with those triggers.

The following listing uses both a SimpleTriggerFactoryBean and a CronTriggerFactoryBean:

The preceding example sets up two triggers, one running every 50 seconds with a starting delay of 10 seconds and one running every morning at 6 AM. To finalize everything, we need to set up the SchedulerFactoryBean, as the following example shows:

More properties are available for the SchedulerFactoryBean, such as the calendars used by the job details, properties to customize Quartz with, and a Spring-provided JDBC DataSource. See the SchedulerFactoryBean javadoc for more information.

As the name implies, you can use @Cacheable to demarcate methods that are cacheable — that is, methods for which the result is stored in the cache so that, on subsequent invocations (with the same arguments), the value in the cache is returned without having to actually invoke the method. In its simplest form, the annotation declaration requires the name of the cache associated with the annotated method, as the following example shows:

In the preceding snippet, the findBook method is associated with the cache named books. Each time the method is called, the cache is checked to see whether the invocation has already been run and does not have to be repeated. While in most cases, only one cache is declared, the annotation lets multiple names be specified so that more than one cache is being used. In this case, each of the caches is checked before invoking the method — if at least one cache is hit, the associated value is returned.

The following example uses @Cacheable on the findBook method with multiple caches:

When the cache needs to be updated without interfering with the method execution, you can use the @CachePut annotation. That is, the method is always invoked and its result is placed into the cache (according to the @CachePut options). It supports the same options as @Cacheable and should be used for cache population rather than method flow optimization. The following example uses the @CachePut annotation:

The cache abstraction allows not just population of a cache store but also eviction. This process is useful for removing stale or unused data from the cache. As opposed to @Cacheable, @CacheEvict demarcates methods that perform cache eviction (that is, methods that act as triggers for removing data from the cache). Similarly to its sibling, @CacheEvict requires specifying one or more caches that are affected by the action, allows a custom cache and key resolution or a condition to be specified, and features an extra parameter (allEntries) that indicates whether a cache-wide eviction needs to be performed rather than just an entry eviction (based on the key). The following example evicts all entries from the books cache:

This option comes in handy when an entire cache region needs to be cleared out. Rather than evicting each entry (which would take a long time, since it is inefficient), all the entries are removed in one operation, as the preceding example shows. Note that the framework ignores any key specified in this scenario as it does not apply (the entire cache is evicted, not only one entry).

You can also indicate whether the eviction should occur after (the default) or before the method is invoked by using the beforeInvocation attribute. The former provides the same semantics as the rest of the annotations: Once the method completes successfully, an action (in this case, eviction) on the cache is run. If the method does not run (as it might be cached) or an exception is thrown, the eviction does not occur. The latter (beforeInvocation=true) causes the eviction to always occur before the method is invoked. This is useful in cases where the eviction does not need to be tied to the method outcome.

Note that void methods can be used with @CacheEvict - as the methods act as a trigger, the return values are ignored (as they do not interact with the cache). This is not the case with @Cacheable which adds data to the cache or updates data in the cache and, thus, requires a result.

Sometimes, multiple annotations of the same type (such as @CacheEvict or @CachePut) need to be specified — for example, because the condition or the key expression is different between different caches. @Caching lets multiple nested @Cacheable, @CachePut, and @CacheEvict annotations be used on the same method. The following example uses two @CacheEvict annotations:

So far, we have seen that caching operations offer many customization options and that you can set these options for each operation. However, some of the customization options can be tedious to configure if they apply to all operations of the class. For instance, specifying the name of the cache to use for every cache operation of the class can be replaced by a single class-level definition. This is where @CacheConfig comes into play. The following examples uses @CacheConfig to set the name of the cache:

@CacheConfig is a class-level annotation that allows sharing the cache names, the custom KeyGenerator, the custom CacheManager, and the custom CacheResolver. Placing this annotation on the class does not turn on any caching operation.

An operation-level customization always overrides a customization set on @CacheConfig. Therefore, this gives three levels of customizations for each cache operation:

It is important to note that even though declaring the cache annotations does not automatically trigger their actions - like many things in Spring, the feature has to be declaratively enabled (which means if you ever suspect caching is to blame, you can disable it by removing only one configuration line rather than all the annotations in your code).

To enable caching annotations add the annotation @EnableCaching to one of your @Configuration classes:

Alternatively, for XML configuration you can use the cache:annotation-driven element:

Both the cache:annotation-driven element and the @EnableCaching annotation let you specify various options that influence the way the caching behavior is added to the application through AOP. The configuration is intentionally similar with that of @Transactional.

The caching abstraction lets you use your own annotations to identify what method triggers cache population or eviction. This is quite handy as a template mechanism, as it eliminates the need to duplicate cache annotation declarations, which is especially useful if the key or condition are specified or if the foreign imports (org.springframework) are not allowed in your code base. Similarly to the rest of the stereotype annotations, you can use @Cacheable, @CachePut, @CacheEvict, and @CacheConfig as meta-annotations (that is, annotations that can annotate other annotations). In the following example, we replace a common @Cacheable declaration with our own custom annotation:

In the preceding example, we have defined our own SlowService annotation, which itself is annotated with @Cacheable. Now we can replace the following code:

The following example shows the custom annotation with which we can replace the preceding code:

Even though @SlowService is not a Spring annotation, the container automatically picks up its declaration at runtime and understands its meaning. Note that, as mentioned earlier, annotation-driven behavior needs to be enabled.

For those who are familiar with Spring’s caching annotations, the following table describes the main differences between the Spring annotations and their JSR-107 counterparts:

JCache has the notion of javax.cache.annotation.CacheResolver, which is identical to the Spring’s CacheResolver interface, except that JCache supports only a single cache. By default, a simple implementation retrieves the cache to use based on the name declared on the annotation. It should be noted that, if no cache name is specified on the annotation, a default is automatically generated. See the javadoc of @CacheResult#cacheName() for more information.

CacheResolver instances are retrieved by a CacheResolverFactory. It is possible to customize the factory for each cache operation, as the following example shows:

Keys are generated by a javax.cache.annotation.CacheKeyGenerator that serves the same purpose as Spring’s KeyGenerator. By default, all method arguments are taken into account, unless at least one parameter is annotated with @CacheKey. This is similar to Spring’s custom key generation declaration. For instance, the following are identical operations, one using Spring’s abstraction and the other using JCache:

You can also specify the CacheKeyResolver on the operation, similar to how you can specify the CacheResolverFactory.

JCache can manage exceptions thrown by annotated methods. This can prevent an update of the cache, but it can also cache the exception as an indicator of the failure instead of calling the method again. Assume that InvalidIsbnNotFoundException is thrown if the structure of the ISBN is invalid. This is a permanent failure (no book could ever be retrieved with such a parameter). The following caches the exception so that further calls with the same, invalid, ISBN throw the cached exception directly instead of invoking the method again:

You do not need to do anything specific to enable the JSR-107 support alongside Spring’s declarative annotation support. Both @EnableCaching and the cache:annotation-driven XML element automatically enable the JCache support if both the JSR-107 API and the spring-context-support module are present in the classpath.

The JDK-based Cache implementation resides under org.springframework.cache.concurrent package. It lets you use ConcurrentHashMap as a backing Cache store. The following example shows how to configure two caches:

The preceding snippet uses the SimpleCacheManager to create a CacheManager for the two nested ConcurrentMapCache instances named default and books. Note that the names are configured directly for each cache.

As the cache is created by the application, it is bound to its lifecycle, making it suitable for basic use cases, tests, or simple applications. The cache scales well and is very fast, but it does not provide any management, persistence capabilities, or eviction contracts.

Ehcache 3.x is fully JSR-107 compliant and no dedicated support is required for it. See JSR-107 Cache for details.

Caffeine is a Java 8 rewrite of Guava’s cache, and its implementation is located in the org.springframework.cache.caffeine package and provides access to several features of Caffeine.

The following example configures a CacheManager that creates the cache on demand:

You can also provide the caches to use explicitly. In that case, only those are made available by the manager. The following example shows how to do so:

The Caffeine CacheManager also supports custom Caffeine and CacheLoader. See the Caffeine documentation for more information about those.

GemFire is a memory-oriented, disk-backed, elastically scalable, continuously available, active (with built-in pattern-based subscription notifications), globally replicated database and provides fully-featured edge caching. For further information on how to use GemFire as a CacheManager (and more), see the Spring Data GemFire reference documentation.

Spring’s caching abstraction can also use JSR-107-compliant caches. The JCache implementation is located in the org.springframework.cache.jcache package.

Again, to use it, you need to declare the appropriate CacheManager. The following example shows how to do so:

Sometimes, when switching environments or doing testing, you might have cache declarations without having an actual backing cache configured. As this is an invalid configuration, an exception is thrown at runtime, since the caching infrastructure is unable to find a suitable store. In situations like this, rather than removing the cache declarations (which can prove tedious), you can wire in a simple dummy cache that performs no caching — that is, it forces the cached methods to be invoked every time. The following example shows how to do so:

The CompositeCacheManager in the preceding chains multiple CacheManager instances and, through the fallbackToNoOpCache flag, adds a no-op cache for all the definitions not handled by the configured cache managers. That is, every cache definition not found in either jdkCache or gemfireCache (configured earlier in the example) is handled by the no-op cache, which does not store any information, causing the target method to be invoked every time.

The jee elements deal with issues related to Jakarta EE (Enterprise Edition) configuration, such as looking up a JNDI object and defining EJB references.

To use the elements in the jee schema, you need to have the following preamble at the top of your Spring XML configuration file. The text in the following snippet references the correct schema so that the elements in the jee namespace are available to you:

The jms elements deal with configuring JMS-related beans, such as Spring’s Message Listener Containers. These elements are detailed in the section of the JMS chapter entitled JMS Namespace Support. See that chapter for full details on this support and the jms elements themselves.

In the interest of completeness, to use the elements in the jms schema, you need to have the following preamble at the top of your Spring XML configuration file. The text in the following snippet references the correct schema so that the elements in the jms namespace are available to you:

This element is detailed in Configuring Annotation-based MBean Export.

You can use the cache elements to enable support for Spring’s @CacheEvict, @CachePut, and @Caching annotations. It it also supports declarative XML-based caching. See Enabling Caching Annotations and Declarative XML-based Caching for details.

To use the elements in the cache schema, you need to have the following preamble at the top of your Spring XML configuration file. The text in the following snippet references the correct schema so that the elements in the cache namespace are available to you: