Answer Posted / b.raja
Connections made via a DataSource object that is
implemented to work with a middle tier connection pool
manager will participate in connection pooling. This can
improve the performance dramatically because creating a new
connection is very expensive.
Connection Pool provides following features:
• Substantial improvement in the performance of DB
application can be accomplished by pre-caching the DB
connection objects
• CPM supplied DB connections are remote enable
• CPM supplied DB connections are cluster aware
• CPM supplied DB connections supports DTM
(distributed TXs)
• CPM supplied DB connections are not actual DB
Connection objects, in turn they are remote object, hence
even though client closes DB connection using con.close()
the actual connection may not be closed instead RMI
connection between client to CPM are closed
• CPM supplied DB connection objects are
serializable, hence client from any where in the network
can access DB connections
The classes and interfaces used for connection pooling are:
1. ConnectionPoolDataSource
2. PooledConnection
3. ConnectionEvent
4. ConnectionEventListener
Connection Pool Manager resided on middle tier
system uses these classes and interfaces behind the scenes.
When the ConnectionPooledDataSource object is called on to
create PooledConnection object, the connection pool manager
will register as a ConnectionEventListener object with the
new PooledConnection object. When the connection is closed
or there is an error, the connection pool manager (being
listener) gets a notification that includes a
ConnectionEvent object.
| Is This Answer Correct ? | 0 Yes | 0 No |
Post New Answer View All Answers
What is metaspace?
What is the difference between RMI and Corba?
Can constructors be synchronized in java?
What is Stream Tokenizer?
Is “abc” a primitive value?
Explain what is orm?
What are the different approaches to represent an inheritance hierarchy?
What is a policy?
Where we can write Rmi registry in the code, without having to write it at the command prompt?
Explain RMI Architecture?
In RMI, inorder to sent the stub reference to the client, is we have to load the server object first into the memory or can we directly sent reference to the client?
Why a component architecture for the java platform?
What is clustering? What are the different algorithms used for clustering?
Do you think that java should have had pointers?
what are memory considerations of jsp compares to other web components?
