Available as of Camel 2.1
The cache component enables you to perform caching operations using EHCache as the Cache Implementation. The cache itself is created on demand or if a cache of that name already exists then it is simply utilized with its original settings.
This component supports producer and event based consumer endpoints.
The Cache consumer is an event based consumer and can be used to listen and respond to specific cache activities. If you need to perform selections from a pre-existing cache, use the processors defined for the cache component.
Maven users will need to add the following dependency to their
pom.xml for this component:
You can append query options to the URI in the following format,
The number of elements that may be stored in the defined cache
The number of elements that may be stored in the defined cache. Options include
- MemoryStoreEvictionPolicy.LFU - Least frequently used
- MemoryStoreEvictionPolicy.LRU - Least recently used
- MemoryStoreEvictionPolicy.FIFO - first in first out, the oldest element by creation time
Specifies whether cache may overflow to disk
Sets whether elements are eternal. If eternal, timeouts are ignored and the
element never expires.
The maximum time between creation time and when an element expires.
Is used only if the element is not eternal
The maximum amount of time between accesses before an element expires
Whether the disk store persists between restarts of the Virtual Machine.
The number of seconds between runs of the disk expiry thread.
Camel 2.8: If you want to use a custom factory which instantiates and creates the EHCache
Type: abstract org.apache.camel.component.cache.CacheManagerFactory
Camel 2.8: Sets a list of EHCache
net.sf.ehcache.event.CacheEventListener for all new caches- no need to define it per cache in EHCache xml config anymore.
Camel 2.8: Sets a list of
org.apache.camel.component.cache.CacheLoaderWrapper that extends EHCache
net.sf.ehcache.loader.CacheLoader for all new caches- no need to define it per cache in EHCache xml config anymore.
Camel 2.10: To configure using a cache key by default. If a key is provided in the message header, then the key from the header takes precedence.
Camel 2.10: To configure using an cache operation by default. If an operation in the message header, then the operation from the header takes precedence.
Camel 2.15: Whether to turn on allowing to store non serializable objects in the cache. If this option is enabled then overflow to disk cannot be enabled as well.
Cache Component options
To use a custom
To use a custom
Camel 2.13/2.12.3: To configure the location of the
ehcache.xml file to use, such as
classpath:com/foo/mycache.xml to load from classpath. If no configuration is given, then the default settings from EHCache is used.
Sending/Receiving Messages to/from the cache
The operation to be performed on the cache. Valid options are
CHECK requires Camel 2.3 onwards.
The cache key used to store the Message in the cache. The cache key is optional if the CACHE_OPERATION is DELETEALL
Header changes in Camel 2.8
The header names and supported values have changed to be prefixed with 'CamelCache' and use mixed case. This makes them easier to identify and keep separate from other headers. The CacheConstants variable names remain unchanged, just their values have been changed. Also, these headers are now removed from the exchange after the cache operation is performed.
The operation to be performed on the cache. The valid options are
The cache key used to store the Message in the cache. The cache key is optional if the CamelCacheOperation is CamelCacheDeleteAll
CamelCacheUpdate operations support additional headers:
Camel 2.11: Time to live in seconds.
Camel 2.11: Time to idle in seconds.
Camel 2.11: Whether the content is eternal.
Sending data to the cache involves the ability to direct payloads in exchanges to be stored in a pre-existing or created-on-demand cache. The mechanics of doing this involve
- setting the Message Exchange Headers shown above.
- ensuring that the Message Exchange Body contains the message directed to the cache
Receiving data from the cache involves the ability of the CacheConsumer to listen on a pre-existing or created-on-demand Cache using an event Listener and receive automatic notifications when any cache activity take place (i.e CamelCacheGet/CamelCacheUpdate/CamelCacheDelete/CamelCacheDeleteAll). Upon such an activity taking place
- an exchange containing Message Exchange Headers and a Message Exchange Body containing the just added/updated payload is placed and sent.
- in case of a CamelCacheDeleteAll operation, the Message Exchange Header CamelCacheKey and the Message Exchange Body are not populated.
There are a set of nice processors with the ability to perform cache lookups and selectively replace payload content at the
Cache Usage Samples
Example 1: Configuring the cache
Example 2: Adding keys to the cache
Example 2: Updating existing keys in a cache
Example 3: Deleting existing keys in a cache
Example 4: Deleting all existing keys in a cache
Example 5: Notifying any changes registering in a Cache to Processors and other Producers
Example 6: Using Processors to selectively replace payload with cache values
Example 7: Getting an entry from the Cache
Example 8: Checking for an entry in the Cache
Note: The CHECK command tests existence of an entry in the cache but doesn't place a message in the body.
Management of EHCache
EHCache has its own statistics and management from JMX.
Here's a snippet on how to expose them via JMX in a Spring application context:
Of course you can do the same thing in straight Java:
You can get cache hits, misses, in-memory hits, disk hits, size stats this way. You can also change CacheConfiguration parameters on the fly.
Cache replication Camel 2.8+
The Camel Cache component is able to distribute a cache across server nodes using several different replication mechanisms including: RMI, JGroups, JMS and Cache Server.
There are two different ways to make it work:
1. You can configure
2. You can configure these three options:
Configuring Camel Cache replication using the first option is a bit of hard work as you have to configure all caches separately. So in a situation when the all names of caches are not known, using
ehcache.xml is not a good idea.
The second option is much better when you want to use many different caches as you do not need to define options per cache. This is because replication options are set per
CacheManager and per
CacheEndpoint. Also it is the only way when cache names are not know at the development phase.
It might be useful to read the EHCache manual to get a better understanding of the Camel Cache replication mechanism.
Example: JMS cache replication
JMS replication is the most powerful and secured replication method. Used together with Camel Cache replication makes it also rather simple.
An example is available on a separate page.