Difference bewteen session.get session.load Hibernate

Note: Before going into the subject, it is advised to know about Eager and Lazy initializations.

In Hibernate, to get the the information of a specific record, we have two methods get() and load(), both defined in Session interface. Then what is the difference between them and when to use which one. This tutorial throws some light over it.

The explanation is more linked to eager and lazy intializations.

Object obj = s.get(Student.class, new Integer(100)); // eager initialization
Object obj1 = s.load(Student.class, new Integer(101)); // lazy initialization

The above statements show the usage of get() and load() methods. A complete program on these methods is available at Hibernate – get, load, delete and Update Methods.

The two methods are defined in org.hibernate.Session interface. First let us see the signature of these methods as defined in Hibernate API.

Object get(Class class, Serializable id) throws HibernateException

Return the persistent instance of the given entity class with the given identifier, or null if there is no such persistent instance. (If the instance, or a proxy for the instance, is already associated with the session, return that instance or proxy.)

Object load(Class class, Serializable id) throws HibernateException

Returns the persistent instance of the given entity class with the given identifier, assuming that the instance exists.

You should not use this method to determine if an instance exists (use get() instead). Use this only to retrieve an instance that you assume exists, where non-existence would be an actual error.

Both methods are used to retrieve only one Java object (record) from the database (for multiple records, we have, list() and iterate() methods).

get()

The get() method returns the Java object from the database as an object of Object class and is required to type cast as follows.

Object obj = s.get(Student.class, new Integer(1));  // where sid is 1
Student s1 = (Student) obj;

We can club the above two statements as follows as one statement.

Student s1 = (Student) s.get(Student.class, new Integer(1));

If the record on the primary key 1 exists, the s1 contains all the column values in the form of object s1. If the record on id 1 does not exist, the s1 contains null value.

Now it is your job to retrieve the values from s1 object as follows using getter methods.

int id = s1.getSid();
String name = s1.getSname();
double marks = s1.getSMarks();
Date joindate = s1.getSjoindate();

All the above is fine and actually what is happening when get() method is called. Hibernate creates a SQL statement as follows and executes.

select * from school where stud_id=1;  

This is called as eager fetching. That is, a call goes to the database to fetch the record.

load()

Student s1 = (Student) s.load(Student.class, new Integer(1));

In the above statement, s1 contains a proxy of Student object and not the actual Student object.

1. What is proxy?

Proxy is a subclass of Student class, like StudentImpl, implemented internally by Hibernate. s1 belongs to such proxy class.

2. Why Hibernate returns a proxy and not actual object? How this is advantageous?

Proxy contains a reference of the primary key id of the actual object. load() is best used to delete a record or update a record. When deletion is required, why the object s1 should contain all the record data which may spread number of columns? To fetch the record, a call should go to the database. When a method call is avoided the performance increases.

3. When the proxy object s1 is converted into a real object of Student?

When you call s1.getName() or s1.getMarks() etc. the object s1 becomes an actual Student object and now database hit is made. This is called as lazy fetching.

4. Why all this round about?

It is very important to know. The whole magic of Hibernate lies here only. s1 proxy object is returned even when the student record by id 1 does not exist also. That is, s1 is never null (incase of get(), s1 can be null also).

Infact, the object s1 returned by the load() method is used for deletion or updation of a record as follows.

s.delete(s1); or s.update(s1); // if id 1 does not exist, exception is raised

When the deletion operation is to be done there is no necessity that s1 should contain the data of all the columns. load() increases the performance for delete and update operations as load() does not hit the database (and hits only when deletion or updation is done). Really, if data is required, get() is preferred.

Special cases

If get(Student.class, new Integer(1)) is called first and later again load(Student.class, new Integer(1)) is called and if the record with primary key 1 exists, the s1 in the both the cases contains a real object. Why? We know get() method returns a Student object and load() returns s proxy object. With get(), a real object exists in the cache memory, the same object is returned to load() also.

If the reverse happens? If load() is called first and then get()? In both cases, s1 contains a proxy object because first load() returns a proxy and remains in the cache memory.

To put it simple:

get( ) load( )
1 Return value null is possible Never returns null
2 Fast if record exists Slow if record exists
3 Used to retrieve object (record) Used for delete etc. operations
4 Eager fetching Lazy fetching
5 Always hits the database Not always hits
6 Does not return proxy object Always returns a proxy object
7 Performance-wise is slow as it may have to make number of rounds to database to get data Better performance. If already in the cache, the record is available to much difference
8 As it returns null if no record exist, the execution continues It throws ObjectNotFoundException, if record not found. Execution terminates if not handled successfully

For the differences between list() and iterate() refer "Difference between Query.list() and Query.iterate() Hibernate".

More about load() method

load() method is overloaded five times in Session interface. One overloaded method is given down for extra discussion.

Object load(Class theClass, Serializable id, LockMode lockmode) throws HibernateException

Return the persistent instance of the given entity class with the given identifier, obtaining the specified lock mode, assuming the instance exists.

Example:

Course c1 = session.load(Course.class, new Integer(100), LockMode.UPGRADE);

Now the object c1 is locked so that other programmers cannot modify the object c1 until all c1 operations are done like calling c1.setDuration(50) method etc as follows.

c1.setDuration(50);
session.flush();
tx.commit();

Other lock modes: WRITE, NONE, READ, UPGRADE_NOWAIT.

Pass your comments to improve the quality of this tutorial session.get session.load.

4 thoughts on “Difference bewteen session.get session.load Hibernate”

  1. I tested. I use get() two times to retrieve the same object with id =1 and I saw the first time hibernate get object information from database, but after that it checking the object information from cache (first and second level of cache are enabled)

  2. You say that get( ) “Always hits the database”, and in special case, when get() is called after load(), get() looks for the object in cache ?

    Will get() always checks the cache before hitting the DB or checks the cache only if the load() is just beforehand ?

Leave a Comment

Your email address will not be published.