I am too confused after reading lot of articles on hibernate caching.

Can anybody please help me in understanding the differences among the following:

1). First level cache
2). Query cache
3). Second level cache

Please don't ask me to read any article. I have already read so many and now confused.

Please help!!!

I have to apply caching in my project.

The sometimes so called "first level cache" is also known as the Session object, that you retrieve from the SessionFactory, sometimes delegating this to the HibernateUtil helper class.

This object is always used in Hibernate. It's the basis. You use it do anything that might have to do with the persistance of your objects.

It is also called first-level cache, because for example, if you ask for an object of the same id twice or more in a short time, it will return the already retrieved object each time, issuing only one query for that.

Obviously, as always with a cache, there're some caveats to be aware of (synchro with db, mostly), so that's why there're some patterns to follow :
The most important is that the session should be short lived and only used for a use case context. Never think using only one session for a whole app session, for example !

Session is not threadsafe, sessionfactory is. So be careful when using sessions objects in a multi-threaded environment.


You should read this, it is only about twenty lines of doc :
http://www.hibernate.org/hib_docs/v3/re ... querycache

As explained, you use it only when you've got queries that have to be executed many times. Obviously not always, because the required overhead to cache the results has to be compensated by the next call performance boost. If you cache queries that are only once or twice called, you see that you're going to get the contrary result of what you want : better perfs.


Second level is to be used for never-modified objects. So you configure this only for some classes (i.e. tables) of your system. It is designed for always accessed objects. So use second level cache only for entities that never change.


One advice : just practice, and then read the parts of the reference documentation when you need it. Don't try to read it in one shot, it's too big to do it. You won't be able to catch up everything at once.

But sorry, you're gonna have to read still a lot to be ok :-)


Well, what do you mean by this ? You have perfs problems ? If not, I do not understand why you're already thinking about caching. Optimization is always to be done after design, never before. :)

Now another question arises here between query cache and second level cache:

1). Can I use second level cache without enabling query cache?

I am asking this because I have written a sample code and tried to use second level cache after enabling it in my cfg.xml file and disabled query cache,

but as per statistics my next call to get the same object is not hitting the cache. From last couple of days, I have tried so many things, but no clue.


2). Is the query cache also available among sessions. I mean whether this also available at session factory level like second level object caching.


Kindly reply

Here's my understanding:

1). Can I use second level cache without enabling query cache?
Yes. I don't use query caching, but I most certainly use the second level cache. In fact, I don't think you can avoid it.

I am asking this because I have written a sample code and tried to use second level cache after enabling it in my cfg.xml file and disabled query cache, but as per statistics my next call to get the same object is not hitting the cache. From last couple of days, I have tried so many things, but no clue.

It depends on how you load the object as to whether or not it will used the cached version or not. For example, using a load() by ID will most certainly use the cache. Conversely some associations don't cache like you'd expect. For example, non-lazy collections won't have entities inside them cached. (I may be oversimplifying here... Corrections anybody?) Also, there are other rules as well. I heartily recommend reading the chapter on caching in Hibernate in Action.

2). Is the query cache also available among sessions. I mean whether this also available at session factory level like second level object caching.
I'm almost certain the query cache is stored in the second level cache mechanism. Check the source code. The documentation talks about cache regions and custom cache expiration policies. That stuff wouldn't make sense if the query cache were only on a per-session basis so it must be per SessionFactory.

Thanks a lot!!!. It clears most of my doubts.

Don't forget the ratings on answers, need those credits!