Improve your overall EJB performance with the use of course grained accessor methods

A common mistake that I often come across in EJB programming is an overuse
of fine-grained accessor methods resulting in an unnecessary resource drain.

In example 1, to create/edit a PersonBean we would be making a separate
request for each of its set methods causing unnecessary amounts of network
traffic,processor utilization, and database resources. Retrieval of the data 
would also require a separate request for each attribute.

Example 1.
public class PersonBean 
	implements SessionBean {

	//member variables
	public long userid;
	public String username;
	public String firstname;
	public String lastname;

	//accessor methods
	public void setUserId(long id) { this.userid = id; }
	public void setUsername	(String username) { this.username = username; }
	public void setFirstname(String firstname) { this.firstname = firstname;  }
	public void setLastname	(String lastname) { this.lastname = lastname; }

	public long getUserId() { return this.userid; }
	public String getUsername() { return this.username; }
	public String getFirstname() { return this.firstname; }
	public String getLastname() { return this.lastname; }

}

Note - In the following example an accessor method that takes all four
attributes as a signature would be adequate but what if PersonBean contained
50 attributes? Things could get sloppy and become a nightmare to debug and
maintain. I only used 4 attributes for this example in an attempt to
maintain simplicity.

A more favorable method that significantly reduces the network consumption 
is to use a data structure object containing all of the necessary variables (example 2). 
We can then instantiate this structure object and pass this to a new setPersonStruct(PersonStruct)
method in PersonBean (example 3). This will reduce the number of requests to
our bean from 4 to 1. We will also need a getPersonStruct() for data
retrieval.

Example 2.
import java.io.Serializable;

public class PersonStruct 
	implements Serializable{

	//member variables
	public long userid;
	public String username;
	public String firstname;
	public String lastname;

	//constructor
	public PersonStruct(long userid, String username, String firstname, String lastname){
		this.userid = userid;
		this.username = username;
		this.firstname = firstname;
		this.lastname = lastname;
	}
}

Example 3.
Create an accessor method for the PersonBean that accepts a PersonStruct as
an argument. We also want an accessor method that returns a instantiated
PersonStruct.

public class PersonBean 
	implements SessionBean {
	//member variables
	public long userid;
	public String username;
	public String firstname;
	public String lastname;

	//new set method taking a PersonStruct as an argument
	public void setPersonStruct(PersonStruct person){
		this.userid = person.userid;
		this.username = person.username;
		this.firstname = person.firstname;
		this.lastname = person.lastname;
	}

	//new get method returning an instantiated PersonStruct
	public PersonStruct getPersonStruct(){
		return new PersonStruct(userid, username, firstname, lastname);
	}

	//accessor methods
	public void setUserId(long id) { this.userid = id; }
	public void setUsername(String username) { this.username = username; }
	public void setFirstname(String firstname) { this.firstname = firstname; }
	public void setLastname(String lastname) { this.lastname = lastname; }
	public long getUserId() { return this.userid; }
	public String getUsername()  { return this.username; }
	public String getFirstname() { return this.firstname; }
	public String getLastname()  { return this.lastname; }
}

Note - It's still a good idea to leave our individual get and set methods intact. If you only wanted to change or retrieve a single
attribute it would make more sense to use the corresponding method for that attribute rather than creating a new structure object.

Ryan Tremaine
rtrema@bellatlantic.net