Caching Mechanism in

Hibernate

1
What is Hibernate

▰ An ORM tool which provides a framework for mapping an object-oriented domain model to a relational database
▰ Simplifies the data creation, data manipulation and data access
▰ Reduces the length of code with increased readability
▰ Generates database independent queries
▰ High performance due to multi layered cache usage

2
Caching

▰ The process of storing copies of data in a temporary storage (cache) for quicker access
▰ Lies between the application and database
▰ Significantly smaller than the main memory or database
▰ Relatively faster and more expensive than the actual data store
▰ Cost efficiency is a trade-off between disk size and speed

3
Caching in Hibernate: At a Glance

▰ A temporary memory that keeps a representation of current database state close to the application
▰ Data is stored in a hydrated (disassembled) state
▰ To Hibernate a cache looks like a map of key/value pairs
▰ Store data in the cache by providing a key and a value
▰ Lookup a value in the cache with a key
▰ Utilizes multi level caching
▻ First level cache (default)
▻ Second level cache (optional)

4
Caching in Hibernate: First Level Cache

▰ Enabled by default
▰ Associates with the Session object
▰ Any object cached in a session will not be visible to other sessions
▰ When the session is closed, all the cached objects will also be lost
▰ An application can use many session object
▰ No DB hit when we’re getting same data in same session
▰ Might get a DB hit when getting same data in different session

5
Caching in Hibernate: Second Level Cache

▰ Not enabled by default

▰ Associates with the Session Factory object
▰ SessionFactory is global for all session objects
▰ Uses a common cache for all the session objects of a session factory
▰ Stores data in the cache as a map of key/value pairs
▰ Can be configured on a per-class and per-collection basis
▰ Can choose among multiple third party cache providers

6
Caching in Hibernate: Representation

7
Caching in Hibernate: Flow Chart

8
The Second Level Cache
Types, Strategies and Configuration of
Hibernate’s Second Level Cache
9
Types of Second Level Cache

▰ Entity Data Cache: The application performs an entity instance lookup by identifier (Primary Key)
▻ Cache key is the identifier value
▻ Cache Value is Entity’s property values
▻ Stores cached entities in a dehydrated (disassembled) format
▻ Assembles the instance when it reads from the cache
▰ Collection Cache: When a collection is initialized
▻ Cache key is the Collection role (i.e.: IndividualPlan[123]#changeForms)
▻ Cache value is a set of identifier values of the collection (set of Change Form Ids)
▰ Natural Identifier Cache: When the application performs an entity instance lookup by a unique key attribute
▻ Cache key is the unique property (i.e: Form ID)
▻ Cached value is the entity instance identifier
10
Types of Second Level Cache

▰ Query Result Cache:

▻ When a JPQL, Criteria or SQL query is executed
▻ Cache key is the query with its params
▻ Cache value is the query result set (may include identifier values)
▻ It’s disabled by default.
▻ Caches only identifiers, not the actual entity state.
▻ Should be used in association with other second level cache

11
Cache Concurrency Strategies (1/4)

▰ READ_WRITE
▻ Caching will work for both read and write
▻ A soft lock is stored in the cache for the entity which is being updated, and released after the commit
▻ Concurrent transactions that access soft-locked entries will fetch the corresponding data directly from database
▻ Maintains read committed isolation
▻ Used for read-mostly data (where it’s critical to prevent stale data in concurrent transactions)
▻ Shouldn’t enable this strategy if data is concurrently modified (by other applications) in the database

12
Cache Concurrency Strategies (2/4)

▰ NONSTRICT_READ_WRITE
▻ Used when its extremely unlikely that two transactions would try to update the same item simultaneously
▻ Cache is updated after a transaction has been committed
▻ No guarantee of consistency between the cache and the database
▻ A transaction may read stale data from the cache
▻ Can configure the duration of the inconsistency window with the expiration policies of your cache provider

13
Cache Concurrency Strategies (3/4)

▰ READ_ONLY
▻ Very suitable for some static reference data that don't change
▻ Throws an exception if an update is triggered
▻ Simplest and best performing strategy
▻ Perfectly safe for use in a cluster

14
Cache Concurrency Strategies (4/4)

▰ TRANSACTIONAL
▻ Available only in environments with a system transaction manager
▻ Guarantees full transactional isolation up to repeatable read
▻ Hibernate does not provide any locking or version checking
▻ Assumes cache provider manages all the transactions
▻ Used for read-mostly data

15
Cache-Provider/Concurrency-Strategy Compatibility

16
Hibernate Cache Best Practices

Caching is useful when:

▰ DB is updated only via Hibernate (Cache is not invalidated if the database is updated in a different way)
▰ Data is updated rarely but it’s read frequently
▰ Noncritical data where slight inconsistency is allowed

Caching is not recommended for:

▰ Critical data that is updated often

▰ Financial data, where decisions must be based on the latest update
▰ Queries with lots of different combinations of parameter values

17
Evicting/Clearing Caches Manually

▰ sessionFactory.getCache().containsEntity(Cat.class, catId): Is this particular Cat currently in the cache

▰ sessionFactory.getCache().evictEntity(Cat.class, catId): Evict a particular Cat
▰ sessionFactory.getCache().evictEntityRegion(Cat.class): Evict all Cats
▰ sessionFactory.getCache().evictEntityRegions(): Evict all entity data
▰ sessionFactory.getCache().containsCollection("Cat.kittens", catId): Is this particular collection currently in the cache
▰ sessionFactory.getCache().evictCollection("Cat.kittens", catId): Evict a particular collection of kittens
▰ sessionFactory.getCache().evictCollectionRegion("Cat.kittens"): Evict all kitten collections
▰ sessionFactory.getCache().evictCollectionRegions(): Evict all collection data

18
Configuring L2 Cache in Hibernate (1/4)

Enabling second level cache and cache provider in persistence.xml

<properties>
...
<property name="hibernate.cache.use_second_level_cache" value="true"/>
<property name="hibernate.cache.region.factory_class"
value="org.hibernate.cache.ehcache.EhCacheRegionFactory"/>
...
</properties>

Making an Entity Cacheable

@Entity
@Table(name = "client")
@Cache(usage = CacheConcurrencyStrategy.READ_WRITE, region = "hibernate-very-big-client")
public class Client {
...
19
Configuring L2 Cache in Hibernate (2/4)

Custom configuration (based on cache names) in hazelcast-hibernate.xml

<map name="hibernate-very-big-*">
<backup-count>0</backup-count>
<time-to-live-seconds>1800</time-to-live-seconds>
<max-idle-seconds>600</max-idle-seconds>
<eviction-policy>LRU</eviction-policy>
<max-size policy="PER_NODE">3000</max-size>
<near-cache>
<time-to-live-seconds>600</time-to-live-seconds>
<max-idle-seconds>300</max-idle-seconds>
<eviction size="1500" eviction-policy="LRU"/>
<invalidate-on-change>true</invalidate-on-change>
</near-cache>
</map>

20
Configuring L2 Cache in Hibernate (3/4)

Custom configuration parameters

▰ Backup Count (0, 1, 2): Number of replicated instances for each data
▰ Time To Live: Maximum time in seconds for each entry to stay in the map
▰ Max Idle Seconds: Maximum time in seconds for each entry to stay idle in the map
▰ Eviction Policy
▻ NONE (Default): No items are evicted and the property Max Size described is ignored
▻ LRU: Least Recently Used.
▻ LFU: Least Frequently Used
▰ Max Size: Max number of key-value pairs stored in the cache

21
Configuring L2 Cache in Hibernate (4/4)

Custom configuration parameters

▰ Near Cache
▻ Cache entries in Hazelcast are partitioned across the cluster members
▻ Retrieving is costly (remote operation) if a key-value pair is owned by another member in your cluster
▻ For mostly-read data, creating local “Near Caches” might be efficient
▻ Increases retrieval speed at the cost of higher memory consumption
▰ Invalidate on Change: Whether the cached entries are evicted when the entries are updated or removed

22
 😁
THANKS!

23