cache.adoc

Cache Abstraction

Since version 3.1, the Spring Framework provides support for transparently adding caching to an existing Spring application. Similar to the transaction support, the caching abstraction allows consistent use of various caching solutions with minimal impact on the code.

In Spring Framework 4.1, the cache abstraction was significantly extended with support for JSR-107 annotations and more customization options.

Understanding the Cache Abstraction

Cache vs Buffer

The terms, “buffer” and “cache,” tend to be used interchangeably. Note, however, that they represent different things. Traditionally, a buffer is used as an intermediate temporary store for data between a fast and a slow entity. As one party would have to wait for the other (which affects performance), the buffer alleviates this by allowing entire blocks of data to move at once rather than in small chunks. The data is written and read only once from the buffer. Furthermore, the buffers are visible to at least one party that is aware of it.

A cache, on the other hand, is, by definition, hidden, and neither party is aware that caching occurs. It also improves performance but does so by letting the same data be read multiple times in a fast fashion.

You can find a further explanation of the differences between a buffer and a cache here.

At its core, the cache abstraction applies caching to Java methods, thus reducing the number of executions based on the information available in the cache. That is, each time a targeted method is invoked, the abstraction applies a caching behavior that checks whether the method has been already invoked for the given arguments. If it has been invoked, the cached result is returned without having to invoke the actual method. If the method has not been invoked, then it is invoked, and the result is cached and returned to the user so that, the next time the method is invoked, the cached result is returned. This way, expensive methods (whether CPU- or IO-bound) can be invoked only once for a given set of parameters and the result reused without having to actually invoke the method again. The caching logic is applied transparently without any interference to the invoker.

Important

This approach works only for methods that are guaranteed to return the same output (result) for a given input (or arguments) no matter how many times they are invoked.

The caching abstraction provides other cache-related operations, such as the ability to update the content of the cache or to remove one or all entries. These are useful if the cache deals with data that can change during the course of the application.

As with other services in the Spring Framework, the caching service is an abstraction (not a cache implementation) and requires the use of actual storage to store the cache data — that is, the abstraction frees you from having to write the caching logic but does not provide the actual data store. This abstraction is materialized by the org.springframework.cache.Cache and org.springframework.cache.CacheManager interfaces.

Spring provides a few implementations of that abstraction: JDK java.util.concurrent.ConcurrentMap based caches, Gemfire cache, Caffeine, and JSR-107 compliant caches (such as Ehcache 3.x). See Plugging-in Different Back-end Caches for more information on plugging in other cache stores and providers.

Important

The caching abstraction has no special handling for multi-threaded and multi-process environments, as such features are handled by the cache implementation.

If you have a multi-process environment (that is, an application deployed on several nodes), you need to configure your cache provider accordingly. Depending on your use cases, a copy of the same data on several nodes can be enough. However, if you change the data during the course of the application, you may need to enable other propagation mechanisms.

Caching a particular item is a direct equivalent of the typical get-if-not-found-then-proceed-and-put-eventually code blocks found with programmatic cache interaction. No locks are applied, and several threads may try to load the same item concurrently. The same applies to eviction. If several threads are trying to update or evict data concurrently, you may use stale data. Certain cache providers offer advanced features in that area. See the documentation of your cache provider for more details.

To use the cache abstraction, you need to take care of two aspects:

• Caching declaration: Identify the methods that need to be cached and their policies.

• Cache configuration: The backing cache where the data is stored and from which it is read.

Declarative Annotation-based Caching

For caching declaration, Spring’s caching abstraction provides a set of Java annotations:

• @Cacheable: Triggers cache population.

• @CacheEvict: Triggers cache eviction.

• @CachePut: Updates the cache without interfering with the method execution.

• @Caching: Regroups multiple cache operations to be applied on a method.

• @CacheConfig: Shares some common cache-related settings at class-level.

The @Cacheable Annotation

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:

@Cacheable("books")
public Book findBook(ISBN isbn) {...}

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.

Note	All the other caches that do not contain the value are also updated, even though the cached method was not actually invoked.

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

@Cacheable({"books", "isbns"})
public Book findBook(ISBN isbn) {...}

Default Key Generation

Since caches are essentially key-value stores, each invocation of a cached method needs to be translated into a suitable key for cache access. The caching abstraction uses a simple KeyGenerator based on the following algorithm:

• If no parameters are given, return SimpleKey.EMPTY.

• If only one parameter is given, return that instance.

• If more than one parameter is given, return a SimpleKey that contains all parameters.

This approach works well for most use-cases, as long as parameters have natural keys and implement valid hashCode() and equals() methods. If that is not the case, you need to change the strategy.

To provide a different default key generator, you need to implement the org.springframework.cache.interceptor.KeyGenerator interface.

Note

The default key generation strategy changed with the release of Spring 4.0. Earlier versions of Spring used a key generation strategy that, for multiple key parameters, considered only the hashCode() of parameters and not equals(). This could cause unexpected key collisions (see SPR-10237 for background). The new SimpleKeyGenerator uses a compound key for such scenarios. If you want to keep using the previous key strategy, you can configure the deprecated org.springframework.cache.interceptor.DefaultKeyGenerator class or create a custom hash-based KeyGenerator implementation.

Custom Key Generation Declaration

Since caching is generic, the target methods are quite likely to have various signatures that cannot be readily mapped on top of the cache structure. This tends to become obvious when the target method has multiple arguments out of which only some are suitable for caching (while the rest are used only by the method logic). Consider the following example:

@Cacheable("books")
public Book findBook(ISBN isbn, boolean checkWarehouse, boolean includeUsed)

At first glance, while the two boolean arguments influence the way the book is found, they are no use for the cache. Furthermore, what if only one of the two is important while the other is not?

For such cases, the @Cacheable annotation lets you specify how the key is generated through its key attribute. You can use SpEL to pick the arguments of interest (or their nested properties), perform operations, or even invoke arbitrary methods without having to write any code or implement any interface. This is the recommended approach over the default generator, since methods tend to be quite different in signatures as the code base grows. While the default strategy might work for some methods, it rarely works for all methods.

The following examples use various SpEL declarations (if you are not familiar with SpEL, do yourself a favor and read Spring Expression Language):

@Cacheable(cacheNames="books", key="#isbn")
public Book findBook(ISBN isbn, boolean checkWarehouse, boolean includeUsed)

@Cacheable(cacheNames="books", key="#isbn.rawNumber")
public Book findBook(ISBN isbn, boolean checkWarehouse, boolean includeUsed)

@Cacheable(cacheNames="books", key="T(someType).hash(#isbn)")
public Book findBook(ISBN isbn, boolean checkWarehouse, boolean includeUsed)

The preceding snippets show how easy it is to select a certain argument, one of its properties, or even an arbitrary (static) method.

If the algorithm responsible for generating the key is too specific or if it needs to be shared, you can define a custom keyGenerator on the operation. To do so, specify the name of the KeyGenerator bean implementation to use, as the following example shows:

@Cacheable(cacheNames="books", keyGenerator="myKeyGenerator")
public Book findBook(ISBN isbn, boolean checkWarehouse, boolean includeUsed)

Note	The key and keyGenerator parameters are mutually exclusive and an operation that specifies both results in an exception.

Default Cache Resolution

The caching abstraction uses a simple CacheResolver that retrieves the caches defined at the operation level by using the configured CacheManager.

To provide a different default cache resolver, you need to implement the org.springframework.cache.interceptor.CacheResolver interface.

Custom Cache Resolution

The default cache resolution fits well for applications that work with a single CacheManager and have no complex cache resolution requirements.

For applications that work with several cache managers, you can set the cacheManager to use for each operation, as the following example shows:

@Cacheable(cacheNames="books", cacheManager="anotherCacheManager") (1)
public Book findBook(ISBN isbn) {...}

1. Specifying anotherCacheManager.

You can also replace the CacheResolver entirely in a fashion similar to that of replacing key generation. The resolution is requested for every cache operation, letting the implementation actually resolve the caches to use based on runtime arguments. The following example shows how to specify a CacheResolver:

@Cacheable(cacheResolver="runtimeCacheResolver") (1)
public Book findBook(ISBN isbn) {...}

1. Specifying the CacheResolver.

Note

Since Spring 4.1, the value attribute of the cache annotations are no longer mandatory, since this particular information can be provided by the CacheResolver regardless of the content of the annotation. Similarly to key and keyGenerator, the cacheManager and cacheResolver parameters are mutually exclusive, and an operation specifying both results in an exception, as a custom CacheManager is ignored by the CacheResolver implementation. This is probably not what you expect.

Synchronized Caching

In a multi-threaded environment, certain operations might be concurrently invoked for the same argument (typically on startup). By default, the cache abstraction does not lock anything, and the same value may be computed several times, defeating the purpose of caching.

For those particular cases, you can use the sync attribute to instruct the underlying cache provider to lock the cache entry while the value is being computed. As a result, only one thread is busy computing the value, while the others are blocked until the entry is updated in the cache. The following example shows how to use the sync attribute:

@Cacheable(cacheNames="foos", sync=true) (1)
public Foo executeExpensiveOperation(String id) {...}

1. Using the sync attribute.

Note	This is an optional feature, and your favorite cache library may not support it. All CacheManager implementations provided by the core framework support it. See the documentation of your cache provider for more details.

Conditional Caching

Sometimes, a method might not be suitable for caching all the time (for example, it might depend on the given arguments). The cache annotations support such use cases through the condition parameter, which takes a SpEL expression that is evaluated to either true or false. If true, the method is cached. If not, it behaves as if the method is not cached (that is, the method is invoked every time no matter what values are in the cache or what arguments are used). For example, the following method is cached only if the argument name has a length shorter than 32:

@Cacheable(cacheNames="book", condition="#name.length() < 32") (1)
public Book findBook(String name)

1. Setting a condition on @Cacheable.

In addition to the condition parameter, you can use the unless parameter to veto the adding of a value to the cache. Unlike condition, unless expressions are evaluated after the method has been invoked. To expand on the previous example, perhaps we only want to cache paperback books, as the following example does:

@Cacheable(cacheNames="book", condition="#name.length() < 32", unless="#result.hardback") (1)
public Book findBook(String name)

1. Using the unless attribute to block hardbacks.

The cache abstraction supports java.util.Optional return types. If an Optional value is present, it will be stored in the associated cache. If an Optional value is not present, null will be stored in the associated cache. #result always refers to the business entity and never a supported wrapper, so the previous example can be rewritten as follows:

@Cacheable(cacheNames="book", condition="#name.length() < 32", unless="#result?.hardback")
public Optional<Book> findBook(String name)

Note that #result still refers to Book and not Optional<Book>. Since it might be null, we use SpEL’s safe navigation operator.

Available Caching SpEL Evaluation Context

Each SpEL expression evaluates against a dedicated context. In addition to the built-in parameters, the framework provides dedicated caching-related metadata, such as the argument names. The following table describes the items made available to the context so that you can use them for key and conditional computations:

Table 1. Cache SpEL available metadata

Name	Location	Description	Example
methodName	Root object	The name of the method being invoked	#root.methodName
method	Root object	The method being invoked	#root.method.name
target	Root object	The target object being invoked	#root.target
targetClass	Root object	The class of the target being invoked	#root.targetClass
args	Root object	The arguments (as array) used for invoking the target	#root.args[0]
caches	Root object	Collection of caches against which the current method is run	#root.caches[0].name
Argument name	Evaluation context	Name of any of the method arguments. If the names are not available (perhaps due to having no debug information), the argument names are also available under the #a<#arg> where #arg stands for the argument index (starting from 0).	#iban or #a0 (you can also use #p0 or #p<#arg> notation as an alias).
result	Evaluation context	The result of the method call (the value to be cached). Only available in unless expressions, cache put expressions (to compute the key), or cache evict expressions (when beforeInvocation is false). For supported wrappers (such as Optional), #result refers to the actual object, not the wrapper.	#result

The @CachePut Annotation

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:

@CachePut(cacheNames="book", key="#isbn")
public Book updateBook(ISBN isbn, BookDescriptor descriptor)

Important

Using @CachePut and @Cacheable annotations on the same method is generally strongly discouraged because they have different behaviors. While the latter causes the method invocation to be skipped by using the cache, the former forces the invocation in order to run a cache update. This leads to unexpected behavior and, with the exception of specific corner-cases (such as annotations having conditions that exclude them from each other), such declarations should be avoided. Note also that such conditions should not rely on the result object (that is, the #result variable), as these are validated up-front to confirm the exclusion.

The @CacheEvict 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:

@CacheEvict(cacheNames="books", allEntries=true) (1)
public void loadBooks(InputStream batch)

1. Using the allEntries attribute to evict all entries from the 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.

The @Caching Annotation

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:

@Caching(evict = { @CacheEvict("primary"), @CacheEvict(cacheNames="secondary", key="#p0") })
public Book importBooks(String deposit, Date date)

The @CacheConfig annotation

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("books") (1)
public class BookRepositoryImpl implements BookRepository {

	@Cacheable
	public Book findBook(ISBN isbn) {...}
}

1. Using @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:

• Globally configured, available for CacheManager, KeyGenerator.

• At the class level, using @CacheConfig.

• At the operation level.

Enabling Caching Annotations

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:

@Configuration
@EnableCaching
public class AppConfig {
}

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

<beans xmlns="http://www.springframework.org/schema/beans"
	xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
	xmlns:cache="http://www.springframework.org/schema/cache"
	xsi:schemaLocation="
		http://www.springframework.org/schema/beans https://www.springframework.org/schema/beans/spring-beans.xsd
		http://www.springframework.org/schema/cache https://www.springframework.org/schema/cache/spring-cache.xsd">

		<cache:annotation-driven/>
</beans>

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.

Note

The default advice mode for processing caching annotations is proxy, which allows for interception of calls through the proxy only. Local calls within the same class cannot get intercepted that way. For a more advanced mode of interception, consider switching to aspectj mode in combination with compile-time or load-time weaving.

Note	For more detail about advanced customizations (using Java configuration) that are required to implement CachingConfigurer, see the {api-spring-framework}/cache/annotation/CachingConfigurer.html[javadoc].

Table 2. Cache annotation settings

XML Attribute	Annotation Attribute	Default	Description
cache-manager	N/A (see the {api-spring-framework}/cache/annotation/CachingConfigurer.html[CachingConfigurer] javadoc)	cacheManager	The name of the cache manager to use. A default CacheResolver is initialized behind the scenes with this cache manager (or cacheManager if not set). For more fine-grained management of the cache resolution, consider setting the 'cache-resolver' attribute.
cache-resolver	N/A (see the {api-spring-framework}/cache/annotation/CachingConfigurer.html[CachingConfigurer] javadoc)	A SimpleCacheResolver using the configured cacheManager.	The bean name of the CacheResolver that is to be used to resolve the backing caches. This attribute is not required and needs to be specified only as an alternative to the 'cache-manager' attribute.
key-generator	N/A (see the {api-spring-framework}/cache/annotation/CachingConfigurer.html[CachingConfigurer] javadoc)	SimpleKeyGenerator	Name of the custom key generator to use.
error-handler	N/A (see the {api-spring-framework}/cache/annotation/CachingConfigurer.html[CachingConfigurer] javadoc)	SimpleCacheErrorHandler	The name of the custom cache error handler to use. By default, any exception thrown during a cache related operation is thrown back at the client.
mode	mode	proxy	The default mode (proxy) processes annotated beans to be proxied by using Spring’s AOP framework (following proxy semantics, as discussed earlier, applying to method calls coming in through the proxy only). The alternative mode (aspectj) instead weaves the affected classes with Spring’s AspectJ caching aspect, modifying the target class byte code to apply to any kind of method call. AspectJ weaving requires spring-aspects.jar in the classpath as well as load-time weaving (or compile-time weaving) enabled. (See Spring configuration for details on how to set up load-time weaving.)
proxy-target-class	proxyTargetClass	false	Applies to proxy mode only. Controls what type of caching proxies are created for classes annotated with the @Cacheable or @CacheEvict annotations. If the proxy-target-class attribute is set to true, class-based proxies are created. If proxy-target-class is false or if the attribute is omitted, standard JDK interface-based proxies are created. (See Proxying Mechanisms for a detailed examination of the different proxy types.)
order	order	Ordered.LOWEST_PRECEDENCE	Defines the order of the cache advice that is applied to beans annotated with @Cacheable or @CacheEvict. (For more information about the rules related to ordering AOP advice, see Advice Ordering.) No specified ordering means that the AOP subsystem determines the order of the advice.

Note

<cache:annotation-driven/> looks for @Cacheable/@CachePut/@CacheEvict/@Caching only on beans in the same application context in which it is defined. This means that, if you put <cache:annotation-driven/> in a WebApplicationContext for a DispatcherServlet, it checks for beans only in your controllers, not your services. See the MVC section for more information.

Method visibility and cache annotations

When you use proxies, you should apply the cache annotations only to methods with public visibility. If you do annotate protected, private, or package-visible methods with these annotations, no error is raised, but the annotated method does not exhibit the configured caching settings. Consider using AspectJ (see the rest of this section) if you need to annotate non-public methods, as it changes the bytecode itself.

Tip

Spring recommends that you only annotate concrete classes (and methods of concrete classes) with the @Cache* annotations, as opposed to annotating interfaces. You certainly can place an @Cache* annotation on an interface (or an interface method), but this works only if you use the proxy mode (mode="proxy"). If you use the weaving-based aspect (mode="aspectj"), the caching settings are not recognized on interface-level declarations by the weaving infrastructure.

Note

In proxy mode (the default), only external method calls coming in through the proxy are intercepted. This means that self-invocation (in effect, a method within the target object that calls another method of the target object) does not lead to actual caching at runtime even if the invoked method is marked with @Cacheable. Consider using the aspectj mode in this case. Also, the proxy must be fully initialized to provide the expected behavior, so you should not rely on this feature in your initialization code (that is, @PostConstruct).

Using Custom Annotations

Custom annotation and AspectJ

This feature works only with the proxy-based approach but can be enabled with a bit of extra effort by using AspectJ.

The spring-aspects module defines an aspect for the standard annotations only. If you have defined your own annotations, you also need to define an aspect for those. Check AnnotationCacheAspect for an example.

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:

@Retention(RetentionPolicy.RUNTIME)
@Target({ElementType.METHOD})
@Cacheable(cacheNames="books", key="#isbn")
public @interface SlowService {
}

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

@Cacheable(cacheNames="books", key="#isbn")
public Book findBook(ISBN isbn, boolean checkWarehouse, boolean includeUsed)

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

@SlowService
public Book findBook(ISBN isbn, boolean checkWarehouse, boolean includeUsed)

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.

JCache (JSR-107) Annotations

Since version 4.1, Spring’s caching abstraction fully supports the JCache standard (JSR-107) annotations: @CacheResult, @CachePut, @CacheRemove, and @CacheRemoveAll as well as the @CacheDefaults, @CacheKey, and @CacheValue companions. You can use these annotations even without migrating your cache store to JSR-107. The internal implementation uses Spring’s caching abstraction and provides default CacheResolver and KeyGenerator implementations that are compliant with the specification. In other words, if you are already using Spring’s caching abstraction, you can switch to these standard annotations without changing your cache storage (or configuration, for that matter).

Feature Summary

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:

Table 3. Spring vs. JSR-107 caching annotations

Spring	JSR-107	Remark
@Cacheable	@CacheResult	Fairly similar. @CacheResult can cache specific exceptions and force the execution of the method regardless of the content of the cache.
@CachePut	@CachePut	While Spring updates the cache with the result of the method invocation, JCache requires that it be passed it as an argument that is annotated with @CacheValue. Due to this difference, JCache allows updating the cache before or after the actual method invocation.
@CacheEvict	@CacheRemove	Fairly similar. @CacheRemove supports conditional eviction when the method invocation results in an exception.
@CacheEvict(allEntries=true)	@CacheRemoveAll	See @CacheRemove.
@CacheConfig	@CacheDefaults	Lets you configure the same concepts, in a similar fashion.

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:

@CacheResult(cacheNames="books", cacheResolverFactory=MyCacheResolverFactory.class) (1)
public Book findBook(ISBN isbn)

1. Customizing the factory for this operation.

Note	For all referenced classes, Spring tries to locate a bean with the given type. If more than one match exists, a new instance is created and can use the regular bean lifecycle callbacks, such as dependency injection.

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:

@Cacheable(cacheNames="books", key="#isbn")
public Book findBook(ISBN isbn, boolean checkWarehouse, boolean includeUsed)

@CacheResult(cacheName="books")
public Book findBook(@CacheKey ISBN isbn, boolean checkWarehouse, boolean includeUsed)

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:

@CacheResult(cacheName="books", exceptionCacheName="failures"
			cachedExceptions = InvalidIsbnNotFoundException.class)
public Book findBook(ISBN isbn)

Enabling JSR-107 Support

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.

Note	Depending on your use case, the choice is basically yours. You can even mix and match services by using the JSR-107 API on some and using Spring’s own annotations on others. However, if these services impact the same caches, you should use a consistent and identical key generation implementation.

Declarative XML-based Caching

If annotations are not an option (perhaps due to having no access to the sources or no external code), you can use XML for declarative caching. So, instead of annotating the methods for caching, you can specify the target method and the caching directives externally (similar to the declarative transaction management advice). The example from the previous section can be translated into the following example:

<!-- the service we want to make cacheable -->
<bean id="bookService" class="x.y.service.DefaultBookService"/>

<!-- cache definitions -->
<cache:advice id="cacheAdvice" cache-manager="cacheManager">
	<cache:caching cache="books">
		<cache:cacheable method="findBook" key="#isbn"/>
		<cache:cache-evict method="loadBooks" all-entries="true"/>
	</cache:caching>
</cache:advice>

<!-- apply the cacheable behavior to all BookService interfaces -->
<aop:config>
	<aop:advisor advice-ref="cacheAdvice" pointcut="execution(* x.y.BookService.*(..))"/>
</aop:config>

<!-- cache manager definition omitted -->

In the preceding configuration, the bookService is made cacheable. The caching semantics to apply are encapsulated in the cache:advice definition, which causes the findBooks method to be used for putting data into the cache and the loadBooks method for evicting data. Both definitions work against the books cache.

The aop:config definition applies the cache advice to the appropriate points in the program by using the AspectJ pointcut expression (more information is available in Aspect Oriented Programming with Spring). In the preceding example, all methods from the BookService are considered and the cache advice is applied to them.

The declarative XML caching supports all of the annotation-based model, so moving between the two should be fairly easy. Furthermore, both can be used inside the same application. The XML-based approach does not touch the target code. However, it is inherently more verbose. When dealing with classes that have overloaded methods that are targeted for caching, identifying the proper methods does take an extra effort, since the method argument is not a good discriminator. In these cases, you can use the AspectJ pointcut to cherry pick the target methods and apply the appropriate caching functionality. However, through XML, it is easier to apply package or group or interface-wide caching (again, due to the AspectJ pointcut) and to create template-like definitions (as we did in the preceding example by defining the target cache through the cache:definitions cache attribute).

Configuring the Cache Storage

The cache abstraction provides several storage integration options. To use them, you need to declare an appropriate CacheManager (an entity that controls and manages Cache instances and that can be used to retrieve these for storage).

JDK ConcurrentMap-based Cache

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:

<!-- simple cache manager -->
<bean id="cacheManager" class="org.springframework.cache.support.SimpleCacheManager">
	<property name="caches">
		<set>
			<bean class="org.springframework.cache.concurrent.ConcurrentMapCacheFactoryBean" p:name="default"/>
			<bean class="org.springframework.cache.concurrent.ConcurrentMapCacheFactoryBean" p:name="books"/>
		</set>
	</property>
</bean>

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-based Cache

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

Caffeine Cache

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:

<bean id="cacheManager"
		class="org.springframework.cache.caffeine.CaffeineCacheManager"/>

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:

<bean id="cacheManager" class="org.springframework.cache.caffeine.CaffeineCacheManager">
	<property name="cacheNames">
		<set>
			<value>default</value>
			<value>books</value>
		</set>
	</property>
</bean>

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

GemFire-based Cache

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 {docs-spring-gemfire}/html/[Spring Data GemFire reference documentation].

JSR-107 Cache

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:

<bean id="cacheManager"
		class="org.springframework.cache.jcache.JCacheCacheManager"
		p:cache-manager-ref="jCacheManager"/>

<!-- JSR-107 cache manager setup  -->
<bean id="jCacheManager" .../>

Dealing with Caches without a Backing Store

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:

<bean id="cacheManager" class="org.springframework.cache.support.CompositeCacheManager">
	<property name="cacheManagers">
		<list>
			<ref bean="jdkCache"/>
			<ref bean="gemfireCache"/>
		</list>
	</property>
	<property name="fallbackToNoOpCache" value="true"/>
</bean>

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.

Plugging-in Different Back-end Caches

Clearly, there are plenty of caching products out there that you can use as a backing store. For those that do not support JSR-107 you need to provide a CacheManager and a Cache implementation. This may sound harder than it is, since, in practice, the classes tend to be simple adapters that map the caching abstraction framework on top of the storage API, as the Caffeine classes do. Most CacheManager classes can use the classes in the org.springframework.cache.support package (such as AbstractCacheManager which takes care of the boiler-plate code, leaving only the actual mapping to be completed).

How can I Set the TTL/TTI/Eviction policy/XXX feature?

Directly through your cache provider. The cache abstraction is an abstraction, not a cache implementation. The solution you use might support various data policies and different topologies that other solutions do not support (for example, the JDK ConcurrentHashMap — exposing that in the cache abstraction would be useless because there would no backing support). Such functionality should be controlled directly through the backing cache (when configuring it) or through its native API.