AR.2. System Overview

AR.2.1. Key Concepts

The purpose of the Jini architecture is to federate groups of devices and software components into a single, dynamic distributed system. The resulting federation provides the simplicity of access, ease of administration, and support for sharing that are provided by a large monolithic system while retaining the flexibility, uniform response, and control provided by a personal computer or workstation.

The architecture of a single Jini system is targeted to the workgroup. Members of the federation are assumed to agree on basic notions of trust, administration, identification, and policy. It is possible to federate Jini systems themselves for larger organizations.

AR.2.1.1. Services

The most important concept within the Jini architecture is that of a service. A service is an entity that can be used by a person, a program, or another service. A service may be a computation, storage, a communication channel to another user, a software filter, a hardware device, or another user. Two examples of services are printing a document and translating from one word-processor format to some other.

Members of a Jini system federate to share access to services. A Jini system should not be thought of as sets of clients and servers, users and programs, or even programs and files. Instead, a Jini system consists of services that can be collected together for the performance of a particular task. Services may make use of other services, and a client of one service may itself be a service with clients of its own. The dynamic nature of a Jini system enables services to be added or withdrawn from a federation at any time according to demand, need, or the changing requirements of the workgroup using the system.

Jini systems provide mechanisms for service construction, lookup, communication, and use in a distributed system. Examples of services include: devices such as printers, displays, or disks; software such as applications or utilities; information such as databases and files; and users of the system.

Services in a Jini system communicate with each other by using a service protocol, which is a set of interfaces written in the Java programming language. The set of such protocols is open ended. The base Jini system defines a small number of such protocols that define critical service interactions.

AR.2.1.2. Lookup Service

Services are found and resolved by a lookup service. The lookup service is the central bootstrapping mechanism for the system and provides the major point of contact between the system and users of the system. In precise terms, a lookup service maps interfaces indicating the functionality provided by a service to sets of objects that implement the service. In addition, descriptive entries associated with a service allow more fine-grained selection of services based on properties understandable to people.

Objects in a lookup service may include other lookup services; this provides hierarchical lookup. Further, a lookup service may contain objects that encapsulate other naming or directory services, providing a way for bridges to be built between a Jini lookup service and other forms of lookup service. Of course, references to a Jini lookup service may be placed in these other naming and directory services, providing a means for clients of those services to gain access to a Jini system.

A service is added to a lookup service by a pair of protocols called discovery and join—first the service locates an appropriate lookup service (by using the discovery protocol), and then it joins it (by using the join protocol).

AR.2.1.3. Java Remote Method Invocation (RMI)

Communication between services can be accomplished using Java Remote Method Invocation (RMI). The infrastructure to support communication between services is not itself a service that is discovered and used but is, rather, a part of the Jini technology infrastructure. RMI provides mechanisms to find, activate, and garbage collect object groups.

Fundamentally, RMI is a Java programming language-enabled extension to traditional remote procedure call mechanisms. RMI allows not only data to be passed from object to object around the network but full objects, including code. Much of the simplicity of the Jini system is enabled by this ability to move code around the network in a form that is encapsulated as an object.

AR.2.1.4. Security

The design of the security model for Jini technology is built on the twin notions of a principal and an access control list. Jini services are accessed on behalf of some entity—the principal—which generally traces back to a particular user of the system. Services themselves may request access to other services based on the identity of the object that implements the service. Whether access to a service is allowed depends on the contents of an access control list that is associated with the object.

AR.2.1.5. Leasing

Access to many of the services in the Jini system environment is lease based. A lease is a grant of guaranteed access over a time period. Each lease is negotiated between the user of the service and the provider of the service as part of the service protocol: A service is requested for some period; access is granted for some period, presumably taking the request period into account. If a lease is not renewed before it is freed—either because the resource is no longer needed, the client or network fails, or the lease is not permitted to be renewed—then both the user and the provider of the resource may conclude that the resource can be freed.

Leases are either exclusive or non-exclusive. Exclusive leases ensure that no one else may take a lease on the resource during the period of the lease; non-exclusive leases allow multiple users to share a resource.

AR.2.1.6. Transactions

A series of operations, either within a single service or spanning multiple services, can be wrapped in a transaction. The Jini transaction interfaces supply a service protocol needed to coordinate a two-phase commit. How transactions are implemented—and indeed, the very semantics of the notion of a transaction—is left up to the service using those interfaces.

AR.2.1.7. Events

The Jini architecture supports distributed events. An object may allow other objects to register interest in events in the object and receive a notification of the occurrence of such an event. This enables distributed event-based programs to be written with a variety of reliability and scalability guarantees.

AR.2.2. Component Overview

The components of the Jini system can be segmented into three categories: infrastructure, programming model, and services. The infrastructure is the set of components that enables building a federated Jini system, while the services are the entities within the federation. The programming model is a set of interfaces that enables the construction of reliable services, including those that are part of the infrastructure and those that join into the federation.

These three categories, though distinct and separable, are entangled to such an extent that the distinction between them can seem blurred. Moreover, it is possible to build systems that have some of the functionality of the Jini system with vari-ants on the categories or without all three of them. But a Jini system gains its full power because it is a system built with the particular infrastructure and programming models described, based on the notion of a service. Decoupling the segments within the architecture allows legacy code to be changed minimally to take part in a Jini system. Nevertheless, the full power of a Jini system will be available only to new services that are constructed using the integrated model.

A Jini system can be seen as a network extension of the infrastructure, programming model, and services that made Java technology successful in the single-machine case. These categories along with the corresponding components in the familiar Java application environment are shown in Figure AR.2.1:

AR.2.2.1. Infrastructure

The Jini technology infrastructure defines the minimal Jini technology core. The infrastructure includes the following:

• A distributed security system, integrated into RMI, that extends the Java platform’s security model to the world of distributed systems.

• The discovery and join protocols, service protocols that allow services (both hardware and software) to discover, become part of, and advertise supplied services to the other members of the federation.

• The lookup service, which serves as a repository of services. Entries in the lookup service are objects written in the Java programming language; these objects can be downloaded as part of a lookup operation and act as local proxies to the service that placed the code into the lookup service.

The discovery and join protocols define the way a service of any kind becomes part of a Jini system; RMI defines the base language within which the Jini technology-enabled services communicate; the distributed security model and its implementation define how entities are identified and how they get the rights to perform actions on their own behalf and on the behalf of others; and the lookup service reflects the current members of the federation and acts as the central marketplace for offering and finding services by members of the federation.

AR.2.2.2. Programming Model

The infrastructure both enables the programming model and makes use of it. Entries in the lookup service are leased, allowing the lookup service to reflect accurately the set of currently available services. When services join or leave a lookup service, events are signaled, and objects that have registered interest in such events get notifications when new services become available or old services cease to be active. The programming model rests on the ability to move code, which is supported by the base infrastructure.

Both the infrastructure and the services that use that infrastructure are computational entities that exist in the physical environment of the Jini system. However, services also constitute a set of interfaces which define communication protocols that can be used by the services and the infrastructure to communicate between themselves.

These interfaces, taken together, make up the distributed extension of the standard Java programming language model that constitutes the Jini programming model. Among the interfaces that make up the Jini programming model are the following:

• The leasing interface, which defines a way of allocating and freeing resources using a renewable, duration-based model

• The event and notification interfaces, which are an extension of the event model used by JavaBeans components to the distributed environment, enable event-based communication between Jini technology-enabled services

• The transaction interfaces, which enable entities to cooperate in such a way that either all of the changes made to the group occur atomically or none of them occur

The lease interface extends the Java programming language model by adding time to the notion of holding a reference to a resource, enabling references to be reclaimed safely in the face of network failures.

The event and notification interfaces extend the standard event models used by JavaBeans components and the Java application environment to the distributed case, enabling events to be handled by third-party objects while making various delivery and timeliness guarantees. The model also recognizes that the delivery of a distributed notification may be delayed.

The transaction interfaces introduce a lightweight, object-oriented protocol enabling applications using Jini technology to coordinate state changes. The transaction protocol provides two steps to coordinate the actions of a group of distributed objects. The first step is called the voting phase, in which each object “votes” whether it has completed its portion of the task and is ready to commit any changes it made. In the second step, a coordinator issues a “commit” request to each object.

The Jini transaction protocol differs from most transaction interfaces in that it does not assume that the transactions occur in a transaction processing system. Such systems define mechanisms and programming requirements that guarantee the correct implementation of a particular transaction semantics. The Jini transaction protocol takes a more traditional object-oriented view, leaving the correct implementation of the desired transaction semantics up to the implementor of the particular objects that are involved in the transaction. The goal of the transaction protocol is to define the interactions that such objects must have to coordinate such groups of operations.

The interfaces that define the Jini programming model are used by the infrastructure components where appropriate and by the initial Jini technology-enabled services. For example, the lookup service makes use of the leasing and event inter faces. Leasing ensures that services registered continue to be available, and events help administrators discover problems and devices that need configuration. The JavaSpaces service, one example of a Jini technology-enabled service, utilizes leasing and events, and also supports the Jini transaction protocol. The transaction manager can be used to coordinate the voting phase of a transaction for those objects that support transaction protocol.

The implementation of a service is not required to use the Jini programming model, but such services need to use that model for their interaction with the Jini technology infrastructure. For example, every service interacts with the Jini lookup service by using the programming model; and whether a service offers resources on a leased basis or not, the service’s registration with the lookup service will be leased and will need to be periodically renewed.

The binding of the programming model to the services and the infrastructure is what makes such a federation a Jini system not just a collection of services and protocols. The combination of infrastructure, service, and programming model, all designed to work together and constructed by using each other, simplifies the overall system and unifies it in a way that makes it easier to understand.

AR.2.2.3. Services

The Jini technology infrastructure and programming model are built to enable services to be offered and found in the network federation. These services make use of the infrastructure to make calls to each other, to discover each other, and to announce their presence to other services and users.

Services appear programmatically as objects written in the Java programming language, perhaps made up of other objects. A service has an interface that defines the operations that can be requested of that service. Some of these interfaces are intended to be used by programs, while others are intended to be run by the receiver so that the service can interact with a user. The type of the service determines the interfaces that make up that service and also define the set of methods that can be used to access the service. A single service may be implemented by using other services.

Example Jini technology-enabled services include the following:

• A printing service, which can print from applications written in the Java programming language as well as legacy applications

• A JavaSpaces service, which can be used for simple communication and for storage of related groups of objects written in the Java programming language

• A transaction manager, which enables groups of objects to participate in the Jini transaction protocol defined by the programming model

AR.2.3. Service Architecture

Services form the interactive basis for a Jini system, both at the programming and user interface levels. The details of the service architecture are best understood once the Jini discovery and Jini lookup protocols are presented.

AR.2.3.1. Discovery and Lookup Protocols

The heart of the Jini system is a trio of protocols called discovery, join, and lookup. A pair of these protocols—discovery and join—occur when a device is plugged in. Discovery occurs when a service is looking for a lookup service with which to register. Join occurs when a service has located a lookup service and wishes to join it. Lookup occurs when a client or user needs to locate and invoke a service described by its interface type (written in the Java programming language) and possibly other attributes. Figure AR.2.2 outlines the discovery process.

Jini discovery/join is the process of adding a service to a Jini system. A service provider is the originator of the service—a device or software, for example. First, the service provider locates a lookup service by multicasting a request on the local network for any lookup services to identify themselves (Figure AR.2.2). Then, a service object for the service is loaded into the lookup service (Figure AR.2.3). This service object contains the Java programming language interface for the service, including the methods that users and applications will invoke to exe cute the service along with any other descriptive attributes.

Services must be able to find a lookup service; however, a service may delegate the task of finding a lookup service to a third party. The service is now ready to be looked up and used, as shown in the following diagram (Figure AR.2.4).

A client locates an appropriate service by its type—that is, by its interface written in the Java programming language—along with descriptive attributes that are used in a user interface for the lookup service. The service object is loaded into the client.

The final stage is to invoke the service, as shown in the following diagram (Figure AR.2.5).

The service object’s methods may implement a private protocol between itself and the original service provider. Different implementations of the same service interface can use completely different interaction protocols.

The ability to move objects and code from the service provider to the lookup service and from there to the client of the service gives the service provider great freedom in the communication patterns between the service and its clients. This code movement also ensures that the service object held by the client and the service for which it is a proxy are always synchronized because the service object is supplied by the service itself. The client knows only that it is dealing with an implementation of an interface written in the Java programming language, so the code that implements the interface can do whatever is needed to provide the service. Because this code came originally from the service itself, the code can take advantage of implementation details of the service that are known only to the code.

The client interacts with a service via a set of interfaces written in the Java programming language. These interfaces define the set of methods that can be used to interact with the service. Programmatic interfaces are identified by the type system of the Java programming language, and services can be found in a lookup service by asking for those that support a particular interface. Finding a service this way ensures that the program looking for the service will know how to use that service, because that use is defined by the set of methods that are defined by the type.

Programmatic interfaces may be implemented either as RMI references to the remote object that implements the service, as a local computation that provides all of the service locally, or as some combination. Such combinations, called smart proxies, implement some of the functions of a service locally and the remainder through remote calls to a centralized implementation of the service.

A user interface can also be stored in the lookup service as an attribute of a registered service. A user interface stored in the lookup service by a Jini technology-enabled service is an implementation that allows the service to be directly manipulated by a user of the system.

In effect, a user interface for a service is a specialized form of the service interface that enables a program, such as a browser, to step out of the way and let the human user interact directly with a service.

In situations in which no lookup service can be found, a client could use a technique called peer lookup instead. In such situations, the client can send out the same identification packet that is used by a lookup service to request service providers to register. Service providers will then attempt to register with the client as though it were a lookup service. The client can select the services it needs from the registration requests it receives in response and drop or refuse the rest.

AR.2.3.2. Service Implementation

Objects that implement a service may be designed to run in a single address space with other, helper, objects especially when there are certain location or security-based requirements. Such objects make up an object group. An object group is guaranteed to always reside in a single address space or virtual machine when those objects are running. Objects that are not in the same object group are isolated from each other, typically by running them in a different virtual machine or address space.

A service may be implemented directly or indirectly by specialized hardware. Such devices can be contacted by the code associated with the interface for the service.

From the service client’s point of view, there is no distinction between services that are implemented by objects on a different machine, services that are downloaded into the local address space, and services that are implemented in hardware. All of these services will appear to be available on the network, will appear to be objects written in the Java programming language, and, only as far as correct functioning is concerned, one kind of implementation could be replaced by another kind of implementation without change or knowledge by the client. (Note that security permissions must be properly granted.)