Using a MarshalledObject to create persistent data


This tutorial should not be the first activation tutorial that you read. This tutorial assumes that you have already read one or more of the three introductory activation tutorials:

In the case of a UnicastRemoteObject, it is easy to pass command-line arguments to the implementation class, because the server program that received those arguments is always running during the lifetime of the remote object implementation. For activatable objects, however, the setup class may exit immediately after registering the activation descriptor with the RMI daemon and registering the stub with the rmiregistry.

The MarshalledObject class provides a flexible mechanism for passing persistence or initialization data through the ActivationDesc, registered with rmid, rather than hard-coding values into the implementation's class file.


Note:  For the remainder of this tutorial, the terms "activatable object implementation", "activatable object," and "implementation" may be used interchangeably to refer to the class, examples.activation.MyPersistentClass, which implements a remote interface and is activatable.

In this tutorial the setup class, examples.activation.Setup4, does two new things:

In this example, if the persistentObjectStore.ser file exists, the activatable object implementation is initialized with the persistent data from the file. Otherwise, if the file does not exist, the activatable object initializes itself as though this is the first time a client has tried to send data.

The client program, examples.activation.Client4, passes a vector of transaction-like data to the activatable object, and that data is added to the implementation object's vector. Each time a client calls the implementation (to add more transaction data), the activatable implementation stores its state (writes the vector) out to the file specified by the MarshalledObject.

This tutorial is organized as follows:

The files needed for this tutorial are:

You may notice that while the client code is included, it is not discussed in a step-by-step manner, like the implementation and setup classes. The reason for this omission, is that the client code for activatable objects is no different than the RMI client code for accessing non-activatable remote objects. Activation is strictly a server-side implementation decision.

For all of the source code used in the activation tutorials, you may choose from these formats:


Creating the remote interface

Create an interface that describes each of the methods that you would like to call remotely. For this example, the remote interface will be examples.activation.YetAnotherRemoteInterface. There are three steps to create a remote interface:

  1. Make the appropriate imports in the interface
  2. Extend java.rmi.Remote
  3. Declare each of the methods that may be called remotely
Step 1:
Make the appropriate imports in your interface
import java.rmi.*;
import java.util.Vector;
Step 2:
Extend java.rmi.Remote
public interface YetAnotherRemoteInterface extends Remote {
Step 3:
Declare each of the methods that may be called remotely
public Vector calltheServer(Vector v) throws RemoteException;

Creating the implementation class

For this example, the implementation class will be examples.activation.MyPersistentClass. There are five steps to create an activatable implementation class that uses a MarshalledObject:

  1. Make the appropriate imports in the implementation class
  2. Extend your class from java.rmi.activation.Activatable
  3. Declare a two-argument constructor in the implementation class
  4. Write the methods that use the MarshalledObject, to save and restore the object's data state
  5. Implement the remote interface method(s)

Step 1:
Make the appropriate imports in the implementation class
import java.io.*; 
import java.rmi.*; 
import java.rmi.activation.*; 
import java.util.Vector; 

Step 2:
Extend your class from java.rmi.activation.Activatable

public class MyPersistentClass extends Activatable
    implements examples.activation.YetAnotherRemoteInterface {
Step 3:
Declare a two-argument constructor in the implementation class

In the constructor, in addition to the normal call to the superclass's constructor, in this example the MarshalledObject is used to specify the file name of the persistent data store. If the file exists, it's used to initialize this object's variable, a Vector named "transactions". If the file object doesn't exist, then the vector is manually initialized. If there is any error reading the file, then object construction fails.

private Vector transactions; 
private File holder; 

public MyPersistentClass(ActivationID id, MarshalledObject data) 
    throws RemoteException, ClassNotFoundException, java.io.IOException { 

    // Register the object with the activation system 
    // then export it on an anonymous port 
    super(id, 0); 

    // Extract the File object from the MarshalledObject that was 
    // passed to the constructor 
    // 
    holder = (File)data.get(); 
    
    if (holder.exists()) { 
        // Use the MarshalledObject to restore my state 
	// 
	this.restoreState(); 
    } else { 
	transactions = new Vector(1,1); 
	transactions.addElement("Initializing transaction vector"); 
    } 

} 

Step 4:
Write the methods that use the MarshalledObject, to save and restore the object's data state

// If the MarshalledObject that was passed to the constructor was 
// a file, then use it to recover the vector of transaction data 
// 
private void restoreState() throws IOException, ClassNotFoundException { 
    File f = holder; 
    FileInputStream fis = new FileInputStream(f); 
    ObjectInputStream ois = new ObjectInputStream(fis); 
    transactions = (Vector)ois.readObject(); 
    ois.close(); 
} 
    
private void saveState() { 
    try { 
	File f  = holder; 
	FileOutputStream fos = new FileOutputStream(f); 
	ObjectOutputStream oos = new ObjectOutputStream(fos); 
	oos.writeObject(getTransactions()); 
	oos.close(); 
    } catch (Exception e) { 
	throw new RuntimeException("Error saving vector of data"); 
    } 
} 

Step 5:
Implement the remote interface method(s)

Add each of the vector elements passed from the client to the object instance and save the updated vector out to a file.

public Vector calltheServer(Vector v) throws RemoteException { 
        
     int limit = v.size(); 
     for (int i = 0; i < limit; i++) { 
         transactions.addElement(v.elementAt(i)); 
     } 
        
     // Save this object's data out to file 
     // 
     this.saveState(); 
     return transactions; 
}

Creating the "setup" class

The job of the "setup" class is to create all the information necessary for the activatable class, without necessarily creating an instance of the remote object. For this example, the setup class will be examples.activation.Setup4.

The setup class passes information about the activatable class to rmid, registers a remote reference (an instance of the activatable class's stub class) and an identifier (name) with the rmiregistry, and then the setup class may exit. There are seven steps to create a setup class:

  1. Make the appropriate imports
  2. Install a SecurityManager
  3. Create an ActivationGroup instance
  4. Create an ActivationDesc instance
  5. Remove the reference to the implementation class creation, and register with rmid
  6. Bind the stub to a name in the rmiregistry
  7. Quit the setup application

Step 1:
Make the appropriate imports in the setup class

import java.io.File; 
import java.rmi.*; 
import java.rmi.activation.*; 
import java.util.Properties; 

Step 2:
Install a SecurityManager

System.setSecurityManager(new RMISecurityManager());

Step 3:
Create an ActivationGroup instance

Note: In this example, for simplicity, we will use a policy file that gives global permission to anyone from anywhere. Do not use this policy file in a production environment. For more information on how to properly open up permissions using a java.security.policy file, please refer to to the following documents:

In the setup application, the job of the activation group descriptor is to provide all the information that rmid will require to contact the appropriate existing JavaTM virtual machine* (JVM) or spawn a new JVM for the activatable object.

Note: In order to run this code on your system, you'll need to change the policy file location to be the absolute path to where you've installed the example policy file that came with the source code.

// Because of the Java 2 security model, a security policy should 
// be specified for the ActivationGroup VM. The first argument
// to the Properties put method, inherited from Hashtable, is 
// the key and the second is the value 
// 
Properties props = new Properties(); 
props.put("java.security.policy", 
   "/home/rmi_tutorial/activation/policy"); 

ActivationGroupDesc.CommandEnvironment ace = null; 
ActivationGroupDesc exampleGroup = new ActivationGroupDesc(props, ace);

// Once the ActivationGroupDesc has been created, register it 
// with the activation system to obtain its ID
//
ActivationGroupID agi = 
   ActivationGroup.getSystem().registerGroup(exampleGroup);

Step 4:
Create an ActivationDesc instance

In the setup application, the job of the activation descriptor is to provide all the information that rmid will require to create a new instance of the implementation class.

Note: In order to run this code on your system, you'll need to change the file URL location to be the location of the directory on your system, where you've installed the example source code.

// Don't forget the trailing slash at the end of the URL
// or your classes won't be found
//	
String location = "file:/home/rmi_tutorial/activation/";

// Pass the file that we want to persist to as the Marshalled
// object
MarshalledObject data =  new MarshalledObject (new File(
    "/home/rmi_tutorial/activation/persistentObjectStore.ser"));

// The second argument to the ActivationDesc constructor will be used 
// to uniquely identify this class; it's location is relative to the  
// URL-formatted String, location.
//
ActivationDesc desc = new ActivationDesc 
    (agi, "examples.activation.MyPersistentClass", location, data);

Step 5:
Declare an instance of your remote interface and register the activation descriptor with  rmid

YetAnotherRemoteInterface yari = 
    (YetAnotherRemoteInterface)Activatable.register(desc);
System.out.println("Got the stub for MyPersistentClass");

Step 6:
Bind the stub, that was returned by the Activatable.register method, to a name in the rmiregistry

Naming.rebind("MyPersistentClass", yari);
System.out.println("Exported MyPersistentClass");

Step 7:
Quit the setup application

System.exit(0);

Compile and run the code

There are six steps to compile and run the code:

  1. Compile the remote interface, implementation, client and setup classes
  2. Run rmic on the implementation class
  3. Start the rmiregistry
  4. Start the activation daemon, rmid
  5. Run the setup program
  6. Run the client

Step 1:
Compile the remote interface, implementation, client and setup classes

% javac -d . YetAnotherRemoteInterface.java 
% javac -d . MyPersistentClass.java 
% javac -d . Client4.java 
% javac -d . Setup4.java 

Step 2:
Run rmic on the implementation class

% rmic -d . examples.activation.MyPersistentClass

Step 3:
Start the rmiregistry

% rmiregistry &

Note:  Before you start the rmiregistry, you must make sure that the shell or window in which you will run the registry, either has no CLASSPATH set or has a CLASSPATH that does not include the path to any classes that you want downloaded to your client, including the stubs for your remote object implementation classes.

If you start the rmiregistry, and it can find your stub classes in its CLASSPATH, it will ignore the server's java.rmi.server.codebase property, and as a result, your client(s) will not be able to download the stub code for your remote object.

Step 4:
Start the activation daemon, rmid

% rmid -J-Djava.security.policy=rmid.policy &
Where rmid.policy is the name of the security policy file for rmid.

Note: By default, rmid now requires a security policy file, that is used to verify whether or not the information in each ActivationGroupDescriptor is allowed to be used to launch a JVM for an activation group. For complete details, please refer to the rmid man page for the Solaris operating environment and the rmid man page for the Microsoft Windows platform.

Step 5:
Run the setup program

Run the setup, setting the codebase property to be the location of the implementation stubs. There are four things that need to go on the same command line:

  1. The "java" command
  2. A property name=value pair that specifies the location of the security policy file
  3. A property to specify where the stub code lives (no spaces from the "-D" all the way though the last "/")
  4. The fully-qualified package name of the setup program.
There should be one space just after the word "java", one between the two properties, and a third one just before the word "examples" (which is very hard to see when you view this as text, in a browser, or on paper).

% java  -Djava.security.policy=/home/rmi_tutorial/activation/policy  -Djava.rmi.server.codebase=file:/home/rmi_tutorial/activation/  examples.activation.Setup4

The codebase property will be resolved to a URL, so it must have the form of "http://aHost/somesource/" or "file:/myDirectory/location/" or, due to the requirements of some operating systems, "file:///myDirectory/location/" (three slashes after the "file:").

While a file: URL is sometimes easier to use for running example code, using the file: URL will mean that the only clients that will be able to access the server are those that can access the same files system as the server (either by virtue of running on the same machine as the server or by using a shared filesystem, such as NFS). If you wish to use an HTTP server, but don't have one available to you, please feel free to download our HTTP server.

Please note that each of these sample URL strings has a trailing "/". The trailing slash is a requirement for the URL set by the java.rmi.server.codebase property, so the implementation can resolve (find) your class definition(s) properly. For more information on setting the java.rmi.server.codebase property from the command line, please take a look at our tutorial on dynamic code downloading using the java.rmi.server.codebase property.

If you forget the trailing slash on the property, or if the class files can't be located at the source (they aren't really being made available for download) or if you misspell the property name, you'll get thrown a java.lang.ClassNotFoundException. This exception will be thrown when you try to bind your remote object to the rmiregistry, or when the first client attempts to access that object's stub. If the latter case occurs, you have another problem as well because the rmiregistry was finding the stubs in its CLASSPATH.

The server output should look like this:

     Got the stub for MyPersistentClass
     Exported MyPersistentClass
Step 6:
Run the client

The argument to the client program is the hostname of the implementation server, in this case, "vector".

% java -Djava.security.policy=/home/rmi_tutorial/activation/policy 
examples.activation.Client4 vector

The first time that the client is run against this implementation, the output should look like this:

	Got a remote reference to the class MyPersistentClass 
        Called the remote method 
	Result: 
	Initializing transaction vector 
	Deposited money 
	Withdrew money 
	Transferred money from Savings 
	Check cleared 
	Point-of-sale charge at grocery store 
The second time that the client is run against this implementation, the output will include five additional "transactions", so it should look like this:
        Got a remote reference to the class MyPersistentClass 
	Called the remote method 
	Result: 
	Initializing transaction vector 
	Deposited money 
	Withdrew money 
	Transferred money from Savings 
	Check cleared 
	Point-of-sale charge at grocery store 
	Deposited money 
	Withdrew money 
	Transferred money from Savings 
	Check cleared 
	Point-of-sale charge at grocery store 
Additionally, you should see the size of the persistentObjectStore.ser file increase, with each subsequent client call. *As used on this web site, the terms "Java virtual machine" or "JVM" mean a virtual machine for the Java platform.
Copyright © 1999 Sun Microsystems, Inc. All Rights Reserved. 
Please send comments to: rmi-comments@java.sun.com 
Sun