CN100512132C

CN100512132C - Method for implementing service logic operation environment

Info

Publication number: CN100512132C
Application number: CNB031094937A
Authority: CN
Inventors: 梁笑凯; 廖江波
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2003-04-10
Filing date: 2003-04-10
Publication date: 2009-07-08
Anticipated expiration: 2023-04-10
Also published as: CN1536818A

Abstract

The invention discloses a method for realizing a kind of service logic running environment, refers to telecommunication service running environment, the method includes: the communication sends the middle event information to dispatcher; the event dispatcher acquires correspondent service example in the service type pool according to the received middle event message, and controls the acquired service example to accomplish the correspondent operation. In the distributed service logic running environment structure, the service example in the one-time conversation process can share the service conversation context generated in the process; the service example can carry on correspondent process to it. The invention provides universal communication platform through message middle component technology. The invention realizes a kind of service logic running environment mechanism without relying on the application program server platform through the message middle component technology, and it can realizes the share of data between distributed service logic running environment joints.

Description

The method that a kind of service logic running environment realizes

Technical field

The present invention relates to telecommunication service running environment field, be meant the method that a kind of service logic running environment realizes especially.

Background technology

JAIN (JAVA API for Intelligent Network) technology is devoted to provide integration application interface (API) and the interface frame based on the JAVA platform, sets up or the integrated unified telecommunication service solution that can cross over diverse network.Wherein, diverse network comprises packet switching network, wireless network and the traditional public phone exchanges network (PSTN) based on Internet protocol (IP) or asynchronous transfer mode (ATM); The telecommunication service solution comprises professional portability and consistency, and external business is visited the characteristics such as fail safe of this integrated network.The JAIN technology can integrate the agreement of the Internet (Internet) and intelligent network (Intelligent Network), becomes new integrated network (IntegratedNetwork).Numerous participants of JAIN technical organization make a series of API and interface frame according to the consistent course of work that defines, and service developer is followed these API and interface frame, can reach the professional purpose of quick develop and field variation.

The work output of JAIN technical organization comprises two main development fields: API protocol specification and API using standard.The API protocol specification is to formulate the standardized signaling protocol interface specification that inserts IP, cable network and wireless network, and signaling protocol comprises: TCAP Transaction Capability Application Part (TCAP), integrated services network User Part (ISUP), intelligent network application protocol (INAP), MAP (MAP), meeting start agreement (SIP) and H.323 wait.The API using standard is interface frame suggestion and an API standard of formulating network security visit (JAIN Parlay), connectivity management, JAIN session/call control (JCC/JCAT) and the professional generation/carrier class service logic of JAIN running environment aspects such as (JSC/SLEE).

Fig. 1 is a JAIN API standard system structural representation, and as shown in Figure 1, JAIN API architecture can be formed by three layers: JAIN application layer 104, JAIN operation layer 110 and JAIN network protocol layer (Protocol Layer) 111.

JAIN network protocol layer 111 support TCAP, ISUP, INAP, MAP, SIP, H.323, agreements such as Media Gateway Control Protocol and Megaco.The effect of JAIN network protocol layer 11l is the API Calls form that is adapted to the JAIN agreement from the particular network message of following real network layer.The real network layer comprises IP112, broadband network (Broadband) 113, cable network (Wireless) 114, PSTN115 and satellite network (Satellite) 116 etc.

JAIN operation layer 110 comprises JAIN service logic running environment (JSLEE, JAIN ServiceLogic Execution Environment) 108 and JAIN session/call control (JCC, JAIN CallControl/JCAT, JAIN Coordination and Transactions) 109.After business development is good, service deployment in JSLEE 108, and professional is handled network event with the pattern of JCC/JCAT 109.JSLEE 108 is integrated piths of a series of JAIN API, is in the position of comparison core in the whole architecture of JAIN API.JSLEE is a JAIN service logic running environment, can be considered as a virtual opetrating system environment of supporting business operation, and its exemplary functions comprises service life cycle is managed, for the operation of business provides various system services etc.

JAIN application layer 104 comprises third party's application (Trusted third-party applications) 103 that the third party of non-trust uses (Untrusted third-partyapplications) 101, JAIN service creation environment (JSCE, JAIN Service CreationEnvironment) 102 and trusts.The third party of non-trust uses 101 and finds and visit professional and each professional ability of disposing by JAIN service provider interface (JSPA, JAIN ServiceProvider Access) 105 in JSLEE108.The third party who trusts uses 103 safe interfaces by standard (Security Interface) 106 and visits the business among the JSLEE 108.JSCE 102 belongs to the JAIN standard, can be independent, and it is the API standard that a cover can be developed JSLEE 108 business fast.

JAIN operation layer 110 communicates with JSPA 105 and safe interface 106 by the telecommunications space 107 of safety.

Fig. 2 simplifies the logic function block diagram for SLEE, as shown in Figure 2, the element of SLEE 201 mainly comprises type of service example pond (Service Instance Pool) 202, event scheduler (ED, Event Dispatcher) 204, system service (SLEE Resource) 205 and data pool 206.Type of service example pond 202 comprises a plurality of srvice instance 203, and required srvice instance 203 is searched and obtained to event scheduler 204 in type of service example pond 202.System service 205 management service life cycles, timer service, alerting service etc.Data pool comprise various data during the storage running database (DataRepository) 207, professional data-business description (Service Description) 209 is described, and the event information (Event Information) 208 that ED 204 may scheduling events is described.The event information (Event Source) 211 that ED204 accepts from the bottom-layer network protocol stack by different network adapter (Network Adaptor) 210.

The specific implementation SLEE that is integrated into of JAIN SLEE API provides an interface frame, its advantage comprises: the API that a series of explications are provided with a kind of standard and reusable mode, can call public function by these API, service supplier is followed these API just can develop new business and new the application fast; Service developer can only be paid close attention to professional realization logic in developing professional process, and does not spend the procotol of being concerned about bottom and move professional ins and outs, and the change shielding of professional and bottom-layer network agreement is come; Service developer can generate new business apace, also can utilize existing service feature to be combined into new business; After business development was good, new business may operate on the interface frame of all and JSLEE compatibility.

Except JAIN SLEE API, the JAIN technical organization also provides the standard that realizes SLEE, specification description professional life cycle model, professional deployment mechanisms etc.Professional deployment mechanisms comprises the structure of professional packing and to the definition of business tine, how to define the deployment descriptor (Deployment Descriptor) that is used to instruct service deployment, service deployment and subscribing interface etc.In addition, also define professional operating mechanism, comprised asynchronous mechanism or the like between operational administrative interface definition, business.

JSLEE architecture and API are realizing that along with existing network technology and software the improvement of technology constantly develops.The advantage of JSLEE also is embodied in its opening, and it does not stipulate concrete realization details, and the api class interface of the interface frame and the standard of an architecture one-level only is provided.Like this, different developers can realize a concrete SLEE according to the technical characterstic of self.Adopt different technology, have the difference on performance and the cost, help the competition and the development of SLEE product on the market, can promote conversely that simultaneously JSLEE architecture and API's is constantly perfect.

Realize a carrier-class SLEE, must consider effective storage, the object of distributed treatment, load balance, database and be connected the ins and outs such as management in type of service example pond.Can realize with component technology from technological layer JAIN SLEE.EJB (Enterprice JAVA BEAN) component technology is exactly the technology of a feasible realization JSLEE, and in addition, JINI, JINI JAVASpace, JES are the component technologys that optionally can realize JSLEE.Some characteristics of JSLEE, for example distributed treatment, load balancing, disposal ability etc., the performance of the apps server platform that adopted of placing one's entire reliance upon.

Component technology is realized if JSLEE adopts EJB, the place one's entire reliance upon performance of apps server platform of the support EJB that adopted of the performance of JSLEE.Though at present some apps server platforms commonly used are declared the performance index that can reach very high, the effect of actual test is unsatisfactory.Some apps server platforms are to EJB processing of request ability the chances are 1600 request/seconds, and along with the increase of client terminal quantity, the response time obviously extends.Though the apps server platform can improve performance by the optimization to system parameters, parameters optimization institute augmented performance does not have change in essence.In addition, some apps server platforms can adopt group machine superimposing technique, though bigger improvement can be arranged the performance of JSLEE,, but be on the basis that increases the software and hardware investment, to realize.Use the EJB component technology to realize that JSLEE need follow EJB1.0 or 2.0 standards fully.Adopt specific apps server platform to do suitable expansion, reach the specific performance properties index or the characteristic that does not relate in some EJB standards is provided with this to the EJB standard.

Also need to follow the development specifications of the apps server platform that is adopted based on some apps server platform development JSLEE, in the performance characteristics scope of apps server platform support, realize JSLEE.Yet for some characteristic, as asynchronous readjustment characteristic, most of apps server platforms are not supported, can only take some flexible methods to realize in design.

Adopt the EJB component technology to realize JSLEE, when disposing JSLEE, process can more complicated.At first need the disposing application program server platform, performance parameter is transferred earlier measure optimization, and then at apps server platform deploy JSLEE.If adopt commercial apps server platform also to need to pay expensive software cost of use and maintenance cost.

Summary of the invention

In view of this, the object of the present invention is to provide a kind of method of service logic running environment realization, realized not relying on the service logic running environment operating mechanism of apps server platform.

In order to achieve the above object, the invention provides the method that a kind of service logic running environment realizes, it is characterized in that, the communication entity of initialization messaging middleware event message, distribute a unique message-oriented middleware address to all communication entities that need receive the message-oriented middleware event message, this method also comprises following steps:

S, execution are used to read the initialization procedure of the initial configuration file of the message-oriented middleware address that comprises event scheduler;

The message-oriented middleware address of the event scheduler that A, communication entity utilization are obtained sends the message-oriented middleware event message to event scheduler;

B, event scheduler are obtained the corresponding business example according to the message-oriented middleware event message of receiving in the type of service pond, and the srvice instance that control is obtained is finished corresponding operating.

Described communication entity is a network protocol stack; Described steps A sends message-oriented middleware service conversation primary event message for the message-oriented middleware address of the event scheduler that the network protocol stack utilization is obtained to event scheduler; Described step B is that event scheduler is according to the message-oriented middleware service conversation primary event message of receiving, in the type of service pond, obtain idle srvice instance, and with this srvice instance activation, service logic running environment node is this srvice instance distribution service instance identification, controls the srvice instance that is obtained then and finishes corresponding operating.

Described communication entity is a network protocol stack; Described steps A sends message-oriented middleware service conversation process event message for the message-oriented middleware address of the event scheduler that the network protocol stack utilization is obtained to event scheduler; Described step B is the message-oriented middleware service conversation process event message that the event scheduler basis is received, obtains the srvice instance of corresponding activation in the type of service pond, and the srvice instance that control is obtained is finished corresponding operating.

Described communication entity is a srvice instance; Described steps A sends message-oriented middleware srvice instance event message for the message-oriented middleware address of the event scheduler that the srvice instance utilization is obtained to event scheduler; Described step B is that event scheduler is according to the message-oriented middleware srvice instance event message of receiving, in the type of service pond, obtain corresponding target service example, event scheduler sends event handling information according to the srvice instance event message to the target service example, and the controlled target srvice instance is finished corresponding operating.

For distributed service logic running environment structure, described step B further comprises: after event scheduler is received event message, judge whether the corresponding business example is positioned on the service logic running environment node at its place, if obtain the corresponding business example in the type of service pond on this service logic running environment node; Otherwise, the service logic running environment node at this event scheduler place is according to the target service logic operation environment node of initial configuration ff to corresponding target service example place, and the message-oriented middleware address that obtains object event scheduler on the target service logic operation environment node, this event scheduler is to the message of object event scheduler transmission message-oriented middleware form, and the object event scheduler obtains the corresponding business example.

For distributed service logic running environment structure, described step B further comprises: the srvice instance place logic operation environment node that is obtained is changed to the main business example with this srvice instance.

Preferably, for distributed service logic running environment structure, the srvice instance that is positioned at same logic operation environment node with the main business example is set to the local service example; This method further comprises following steps:

D, local service example call data access interface;

E, data access interface be by calling local data access tool access service session context data, and the service conversation context data is operated;

F, local data access tool will return to the local service example to the operating result of service conversation context data.

Described service conversation context data is stored in the service logic running environment node at main business example place.

Preferably, for distributed service logic running environment structure, the srvice instance that is positioned at Different Logic running environment node with the main business example is set to the teleaction service example; This method further may further comprise the steps:

D1, teleaction service example call data access interface;

E1, data access interface call the remote data access instrument, the remote data access instrument sends message-oriented middleware message to the teledata agency of main business example place service logic running environment node, after the teledata agency receives message-oriented middleware message, call local data access tool access service session context data, and the service conversation context data is operated;

F1, local data access tool will return to described teledata agency to the operating result of service conversation context data, described teledata agency will return to described remote data access instrument to the operating result of service conversation context data, and described remote data access instrument will return to the teleaction service example to the operating result of service conversation context data.

In initialization procedure, generate described local data access tool and remote data access instrument.

For distributed service logic running environment structure, described steps A further comprises: the event scheduler of network protocol stack on any service logic running environment node sends service conversation primary event message.

The present invention has realized not relying on the SLEE operating mechanism of apps server platform by the message-oriented middleware technology.The present invention provides general communication platform by the message-oriented middleware technology for SLEE.And, in distributed SLEE structure, which SLEE node the service deployment person can take all factors into consideration service deployment in according to conditions such as load balancing, geographical position, has realized flexible purpose of disposing business, and can realize the data sharing of distributed SLEE by the message-oriented middleware technology.

Description of drawings

Fig. 1 is a JAIN API architectural schematic;

Fig. 2 simplifies the logic function block diagram for SLEE;

Fig. 3 is non-distributed SLEE structural representation;

Fig. 4 is distributed SLEE structural representation;

Fig. 5 is distributed SLEE Session Context data sharing schematic diagram.

Embodiment

For making the purpose, technical solutions and advantages of the present invention clearer, the present invention is described in further detail below in conjunction with accompanying drawing.

The present invention has introduced message-oriented middleware (MOM, Message OrientedMiddleware) technology in SLEE, realize the mutual of information between communication entity by the MOM technology, at first the MOM technology is introduced below.

The MOM technology is used for providing the communications platform of an efficient general between communication object.MOM technical support synchronization message and asynchronous message.Send synchronization message, sender's process can get clogged, and up to having received the message return value, just can continue to carry out follow-up work.Send asynchronous message and do not have obstructing problem, after the sender sends message, can handle follow-up flow process.The communication entity object that each need receive MOM message all needs to distribute a pair of unique module id (MID) and process identification (PID) (PID).MID represents physically independently module No.; PID represents process number, can be used for needing to receive in the identification application specific program module of MOM message.The MOM message sending end will be sent to MOM message the destinations traffic entity, must know the MID and the PID of destinations traffic entity.A pair of unique MID and PID have constituted the address that MOM message sends, and also can be described as the MOM address.

The receiving terminal communication entity of MOM message need be realized the function of two specific JAVA interfaces, is respectively applied for the processing procedure of receiving after asynchronous message and the synchronization message.After MOM receiving terminal communication entity is received MOM message, can call in these two interfaces one respectively according to the MOM type of message, realize the triggering of MOM message event.

SLEE can be divided into non-distributed SLEE and distributed SLEE from its running environment.Wherein, non-distributed SLEE can only run on the physical node, i.e. a JAVA virtual machine (JVM).For business, all business all are deployed on the same physical node, and the communication between business is all carried out on same JVM; Distributed SLEE can only run on the different physical nodes, on the promptly different JVM.

Fig. 3 is non-distributed SLEE structural representation, and as shown in Figure 3, network protocol stack 302 operates on the physical node or JVM different with SLEE 301.307 pairs of a plurality of type of service examples of Service Manager pond 202 manages.Be sent to the ED 204 of SLEE 301 from the internet message 306 of network protocol stack 302 with the form of MOM message, ED 204 carries out subsequent treatment according to the type identification of this incident.

For obtaining the system processing power of better extensibility and Geng Gao, SLEE must support distributed operating mechanism, and system loading can be shared by a plurality of physical nodes.For business, service deployment is on different physical nodes, and the communication between business may be carried out on different JVM.

Fig. 4 is distributed SLEE structural representation, and as shown in Figure 4, each SLEE node all has an ED 204.By ED 204, can intercom mutually between distributed SLEE node, for example, SLEE node A 401 and SLEE Node B 402.Network protocol stack A 403 is sent to the ED 204 of certain SLEE node with the form of MOM message with internet message 306, the ED 204 of SLEE node A 401 for example, and ED 204 handles according to the type identification of incident.Simultaneously ED 204 in the future the internet message 306 of automatic network protocol stack A 403 be forwarded to the ED 204 of other SLEE nodes, for example ED 204 of SLEE Node B 402.

Srvice instance A203 among the SLEE node A 401 can call 406 srvice instance that go to call in the SLEE Node B 402 by distribution.Call about the distribution between srvice instance, can be described below.

Whole distributed SLEE platform can be made up of a plurality of SLEE nodes that operate in different physical locations, and business can be deployed on any one SLEE node.Professional specifically be deployed on which SLEE node can be according to the load balancing of each SLEE node, or the conditions such as convenience in geographical position are taken all factors into consideration.

In addition, no matter the distributed SLEE structure of right and wrong still is distributed SLEE structure, all srvice instance all can be directly and network protocol stack communicate MOM message of srvice instance structure, by MOM message is sent to the objective network protocol stack, make the objective network protocol stack carry out corresponding operating.

All communication entities that need receive MOM message all need to distribute a unique address designation, and promptly a pair of unique MID and the combination of PID are as the destination address of transmitting terminal transmission MOM message.For non-distributed SLEE structure, needing to distribute the communication entity of MID and PID is ED.In addition, the network protocol stack that need communicate also needs to distribute MID and PID for it.

In non-distributed SLEE structure, SLEE can dispose an initial configuration file, has comprised the MID and the PID that distribute to its ED in the initial configuration file, and the MID and the PID that distribute to network protocol stack.

In non-distributed SLEE structure, if service conversation primary event (Service InitialEvent), carry professional type identification in the service conversation primary event message that ED receives, ED searches and obtains the srvice instance of a free time in the type of service example pond of respective type by Service Manager.SLEE activates idle srvice instance and it is carried out initial configuration, disposes some initial parameters.SLEE is srvice instance distribution service instance identification (Instance ID) simultaneously.If current service conversation does not also have distribution service session identification (Session ID), then SLEE distributes Session ID for the current business session, and the incident in ensuing processing procedure and the service conversation process described below is basic identical.

If the incident in the service conversation process, ED is according to Session ID that carries in the service conversation process event message and Instance ID, finds srvice instance in the corresponding activation by Service Manager in type of service example pond.Srvice instance may communicate with other srvice instance in the same service conversation process in the process of processing events, and perhaps the network protocol stack with bottom communicates.When above-mentioned srvice instance and other srvice instance communicated, srvice instance was sent to ED with the srvice instance event message.ED is sent to the corresponding business example according to the srvice instance event message with event handling information, makes this srvice instance go to handle events corresponding.

In distributed SLEE structure, the MOM technology is similarly the communications platform that communication entity provides an efficient general, communication entity is that mutual communication is carried out on the basis with this communications platform, and the communication in all distributed SLEE nodes between communication entity is based on all that MOM message carries out.Wherein synchronous MOM message is adopted in the communication between the ED, can simplify the design of ED like this.The MOM Message Processing of ED adopts multi-thread mechanism, and ED need safeguard a thread resources pond, and promptly ED whenever receives a MOM message, just takes out an idle thread and activate this thread from the thread resources pond, by moving this thread process MOM message.Adopt synchronous MOM message between the ED, transmitting terminal ED is after sending a MOM message, and in the response process of waiting for this MOM message, just this thread is in blocked state, and other threads can other MOM message of parallel processing.On macroscopic view, ED can many MOM message of parallel processing, and the dispatching efficiency of ED and throughput only depend on that JVM is to the dispatching efficiency of thread and the size in thread resources pond.

In the process of disposing distributed SLEE node, distribute a unique SLEE identifier (SLEE ID) in order to identify this SLEE node for each SLEE node.Each network protocol stack also can distribute a unique identifier (ID).

All communication entities that need receive MOM message all need to distribute a unique address designation, and promptly a pair of unique MID and the combination of PID are as the destination address of transmitting terminal transmission MOM message.For distributed SLEE node, needing to distribute the communication entity of MID and PID is the ED of each SLEE node.In addition, all network protocol stacks that need communicate all need to distribute MID and PID.

In distributed SLEE structure, each SLEE node all can dispose an initial configuration file, the SLEE ID that distributes to this SLEE node and MID and the PID of ED thereof have been comprised in the initial configuration file, the SLEE ID of other all SLEE nodes reaches MID and the PID of ED corresponding with it, distributes to ID and the MID and the PID of all-network protocol stack.

When each SLEE node starts initialization procedure, can remove to read the initial configuration file, MID and the PID of SLEE ID, its ED of self SLEE node are read in internal memory, the communication entity of initialization transmitting-receiving MOM message, as dispose some initial parameters etc., and the MID of communication entity self and PID be tied on the MOM communication entity, communication entity just can receive MOM message like this.Like this, communication entity just can receive the message from the MOM of other communication entities.Simultaneously, the SLEE initialization procedure also reads in internal memory to the MID of the SLEE ID of other SLEE nodes and corresponding with it ED and PID, the ID of all-network protocol stack and MID and the PID that distributes for the all-network protocol stack, and sets up the ID of the MID of ED of SLEE ID and this SLEE and PID, network protocol stack respectively and be the MID of its distribution and PID corresponding relation each other.Like this, just can find the MOM address that communicates with the respective communication entity by ID.

Set up a service conversation (Session) process and be that the ED of certain SLEE node receives that the service conversation primary event message that network protocol stack sends begins from the SLEE platform, carry professional type identification in the service conversation primary event message.Network protocol stack is sent to any one SLEE node to service conversation primary event message randomly.

After the SLEE node is received service conversation primary event message, if this type of service is deployed on the SLEE node that receives service conversation primary event message, the process that the service conversation primary event is set up in following processing procedure and the foregoing non-distributed SLEE structure is basic identical, needs this srvice instance to be set to the main business example of current business session simultaneously.

If this type of service is not deployed on the SLEE node at the ED place that receives service conversation primary event message, when each SLEE node has all been safeguarded all types of service of deployment, the SLEE ID relation table of the SLEE node at each type of service place, this SLEE node can find the SLEE ID of the SLEE node at present type of service place, find the MOM address of its ED, the i.e. MID of this ED and PID by the SLEE ID of SLEE node.The ED that receives the SLEE node of service conversation primary event message constructs a MOM message that is used for communicating by letter between ED, and service conversation primary event message is encapsulated in the MOM message, is sent to the ED of the target SLEE node of having disposed this type of service.After the ED of target SLEE node received this service conversation primary event message, its processing procedure and the above process were basic identical.

Session ID comprises SLEE ID and Call ID, wherein SLEE ID is the sign of the SLEE node of this service conversation process of initiation, Call ID is used to organize data for this sign of calling out (distribution between this srvice instance is called), realizes the service conversation process.

Instance ID comprises SLEE ID, Self ID and ServiceKey, and SLEE ID is the sign of the SLEE node that this srvice instance moved, and Self ID is the sign of different business example, and ServiceKey is the sign of type of service under this srvice instance.

If the incident in the service conversation process, ED judges according to the Session ID that carries in the service conversation process event message whether this type of service is deployed on the SLEE node at its place, if ED finds srvice instance in the corresponding activation by Service Manager in type of service example pond; Otherwise the SLEE node at ED place finds the MOM address of its ED, the i.e. MID of this ED and PID according to the SLEE ID that comprises among the Session ID.The ED that receives the SLEE node of service conversation process event message constructs a MOM message that is used for communicating by letter between ED, and service conversation process event message is encapsulated in the MOM message, is sent to the ED of the target SLEE node of having disposed this type of service.After the ED of target SLEE node received this service conversation process event message, its processing procedure and the above process were basic identical.

When the srvice instance of srvice instance and other types of service communicated, srvice instance was sent to ED with the srvice instance event message.ED judges according to the Session ID that carries in the srvice instance event message whether this type of service is deployed on the SLEE node at its place, if, ED is sent to the corresponding business example according to the srvice instance event message with event handling information, makes this srvice instance go to handle events corresponding; Otherwise the SLEE node at ED place finds the MOM address of its ED, the i.e. MID of this ED and PID according to the SLEE ID that comprises among the Session ID.The ED that receives the SLEE node of srvice instance event message constructs a MOM message that is used for communicating by letter between ED, and the srvice instance event message is encapsulated in the MOM message, is sent to the ED of the target SLEE node of having disposed this type of service.After the ED of target SLEE node received this srvice instance event message, its processing procedure and the above process were basic identical.

Realize that the problem that distributed SLEE need consider has: shield physical location between business, network protocol stack shields professional physical location and data sharing.

The shielding physical location is meant when professional A calls professional B between business, and professional A does not need the physical location of knowing that professional B is concrete, and only the interface that need call mutually is unified getting final product.

Network protocol stack shields professional physical location and is meant that network protocol stack passes through to send the message initiated srvice instance of service conversation primary event to the SLEE node, network protocol stack does not need the physical location of knowing that this srvice instance is concrete, only need service conversation primary event message is sent to any one SLEE node, transmit service conversation primary event message by ED and get final product.

A service conversation process can comprise the srvice instance of a plurality of other types of service except the main business example, the situation that in the service conversation process, can occur calling mutually between srvice instance, and the distribution that Here it is between srvice instance is called.It is that ED by distributed SLEE node transmits MOM message and realizes that distribution between srvice instance is called.The srvice instance that initiation is called is sent to the srvice instance event message ED of its place node.ED is sent to the target service example according to the srvice instance event message with event handling information, makes this srvice instance carry out associative operation.

The SLEE node at the srvice instance place that initiation is called distributes Event ID for this calls, and invoked srvice instance is known associative operation by Event ID.Front end in message has Event ID, Session ID, initiates called side srvice instance sign (Source Instance ID) and the target service instance identification that is called (Destination Instance ID).The srvice instance sign of Source Instance ID for initiating to call between srvice instance, Destination Instance ID is invoked target service instance identification.But Session ID and Instance ID are the classes of serializing, can transmit on network.

When ED manages business distribution between example when calling, when ED manages business the MOM message that distribution is called between example according to the SLEE ID among the Destination Instance ID, if SLEE ID is exactly the SLEE ID of self, then invoked srvice instance just operates on this SLEE node, ED only need find the corresponding business example, and the event handling interface that calls this srvice instance gets final product; If SLEEID is not the SLEE ID of self, ED can be according to the initial configuration ff of its place SLEE node SLEE ID to the SLEE node at corresponding service example place.ED constructs a MOM message that is used for communicating by letter between ED, and the distribution message call between srvice instance is encapsulated in this MOM message, is sent to the ED of target SLEE ID by synchronous MOM message, realizes that the distribution between srvice instance is called.After the ED of target SLEE ID received this MOM message, treatment step and the internodal srvice instance invoked procedure of above-mentioned same SLEE were basic identical.

A service conversation (calling) process can generate lot of data, being referred to as these data is service conversation context (Session Context) data, be used for the data of centralized stores service conversation process about attribute and parameter, these SessionContext data can be visited and revise to the srvice instance that participates in this time service conversation.Because srvice instance can be more frequent to the visit and the modification of Session Context data, be to guarantee access efficiency and Session Context safety of data and integrality, adopt centralized system that Session Context data are stored and managed.The SLEE node that the position of storage is moved for the main business example (Main Instance) of initiating a service conversation process.For a plurality of srvice instance that participate in a service conversation, they geographically can distribute, so the mode that they visit corresponding Session Context data also distributes.But the physical location of Session Context deposit data should all be transparent to all srvice instance, that is to say, the interface of srvice instance visit Session Context data should be unified, the physical location that is Session Context data is conductively-closed to srvice instance, realizes sharing of Session Context data with this.

For a large amount of Session Context data that generate in the service conversation process, the srvice instance that participates in this service conversation can carry out frequent access and modification in the service conversation process.For integrality and the fail safe that guarantees data, adopt centralized system that these Session Context data are managed and store, the position of storage is the SLEE node that the main business example of a service conversation process of initiation is moved, the SLEE node that the SLEE ID that promptly comprises in Session ID is identified.Because participating in a plurality of srvice instance of a service conversation process may be distributed on the different SLEE nodes, being necessary for all srvice instance provides the unified interface of visit Session Context data, and making Session Context data storage location is conductively-closeds when srvice instance visit Session Context data.

When srvice instance visit Session Context data, srvice instance can be positioned on the different SLEE nodes with the main business example, therefore will be called the local service example with the srvice instance that the main business example is positioned at same SLEE node, accordingly, be called the teleaction service example with the srvice instance that the main business example is positioned at different SLEE nodes.

Fig. 5 is distributed SLEE Session Context data sharing schematic diagram, and as shown in Figure 5, if srvice instance A501 is the main business example, then Session Context storage is in the distributed SLEE node A401 at main business example 501 places.Srvice instance A501 and srvice instance B502 are the different business example that belongs to a conversation procedure.Demand at all srvice instance visit SessionContext data, a unified Session Context data access interface (DataAccess) 503 is provided, defined all operations of srvice instance visit Session Context data in the data access interface 503, for example visited, operation such as modification.Data access interface 503 provides two different classes to realize data access, one is local data access tool (LocalDataAccessImp1) class 504 that is used to visit local Session Context data, and another is remote data access instrument (RemoteDataAccessImp1) class 505 that is used to visit long-range SessionContext data.SLEE node A401 also needs to realize teledata agency (RemoteDataAgent) class 506, such is as the agency of the Session Context data that are positioned at the teleaction service example B502 visit SLEE node A 401 on the SLEE Node B 402, it can the receiving remote data access synchronous MOM message, and resolve MOM message, the MOM message conversion of each remote data access is become different calling.Have only LocalDataAccessImp1 class 504 can visit local Session Context data center, therefore RemoteDataAgent class 506 can generate quoting of a LocalDataAccessImp1 class 508 at initialization procedure, by calling the local Session Context of the operational access data in the corresponding LocalDataAccessImp1 class 508, and the return results by synchronous MOM message is to the teleaction service example of initiating the SessionContext data access calls.

RemoteDataAgent class in each SLEE node all needs to receive MOM message, therefore also needs to distribute MOM address, promptly a pair of unique MID and PID for the RemoteDataAgent class.This MOM address also is configured in initial configuration file in SLEE node initializing process, and the MOM address of each RemoteDataAgent class is carried out related with its affiliated SLEE ID.The SLEE node reads and stores the MOM address of the RemoteDataAgent class of all SLEE nodes from initial configuration file when starting.

The realization of LocalDataAccessImp1 class can directly go to visit Session Context data center by the mode of calling local API.

Realization for the RemoteDataAgent class can be encapsulated into action type and entrained parameter in the MOM message by to the different MOM message of each data access interface that need realize structure.The RemoteDataAccessImp1 class is according to the srvice instance Instance ID that calls data access interface, can know the Session ID that srvice instance is affiliated, can know the SLEE ID of the SLEE node at SessionContext data place by Session ID, can know the MOM address of RemoteDataAgent class by the SLEE ID of this SLEE node, send synchronous MOM message to this MOM address, the return results of resolving synchronous MOM message is exactly the result that this visit Session Context data obtain.

In initialization procedure, DataAccess is corresponding with corresponding class, if this DataAccess and main business example are then corresponding with local class with DataAccess on same SLEE node; If DataAccess and main business example are then corresponding with long-range class with DataAccess not on same SLEE node.DataAccess provides unified interface for srvice instance visit Session Context data, and making the memory location of Session Context data is transparent for the srvice instance of visiting it.

During local service instance access Session Context data, because DataAccess is corresponding with local class, so DataAccess calls the LocalDataAccessImp1 class automatically, the LocalDataAccessImp1 class is visited Session Context data and it is carried out associative operation, then operating result is returned to this local service example.

During teleaction service instance access Session Context data, because DataAccess is corresponding with long-range class, so DataAccess calls the RemoteDataAccessImp1 class automatically, this RemoteDataAccessImp1 class sends synchronous MOM message according to the initial configuration file to the RemoteDataAgent class of main business example place SLEE node, carry action type and the pairing parameter of this action type in this MOM message to the SessionContext data, after the RemoteDataAgent class is received MOM message, MOM message is resolved, call the LocalDataAccessImp1 class that generates in its initialization procedure then, this LocalDataAccessImp1 class is visited Session Context data and it is carried out associative operation, then operating result is returned to the RemoteDataAgent class, the RemoteDataAgent class returns to the RemoteDataAccessImp1 class of teleaction service example place SLEE node with operating result by synchronous MOM message, and last RemoteDataAccessImp1 class returns to the teleaction service example with operating result.

In a word, the above is preferred embodiment of the present invention only, is not to be used to limit protection scope of the present invention.

Claims

1, a kind of method of service logic running environment realization, it is characterized in that, the communication entity of initialization messaging middleware event message distributes a unique message-oriented middleware address to all communication entities that need receive the message-oriented middleware event message, and this method also comprises following steps:

2, method according to claim 1 is characterized in that: described communication entity is a network protocol stack;

Described steps A sends message-oriented middleware service conversation primary event message for the message-oriented middleware address of the event scheduler that the network protocol stack utilization is obtained to event scheduler;

Described step B is that event scheduler is according to the message-oriented middleware service conversation primary event message of receiving, in the type of service pond, obtain idle srvice instance, and with this srvice instance activation, service logic running environment node is this srvice instance distribution service instance identification, controls the srvice instance that is obtained then and finishes corresponding operating.

3, method according to claim 1 is characterized in that: described communication entity is a network protocol stack;

Described steps A sends message-oriented middleware service conversation process event message for the message-oriented middleware address of the event scheduler that the network protocol stack utilization is obtained to event scheduler;

Described step B is the message-oriented middleware service conversation process event message that the event scheduler basis is received, obtains the srvice instance of corresponding activation in the type of service pond, and the srvice instance that control is obtained is finished corresponding operating.

4, method according to claim 1 is characterized in that: described communication entity is a srvice instance;

Described steps A sends message-oriented middleware srvice instance event message for the message-oriented middleware address of the event scheduler that the srvice instance utilization is obtained to event scheduler;

Described step B is that event scheduler is according to the message-oriented middleware srvice instance event message of receiving, in the type of service pond, obtain corresponding target service example, event scheduler sends event handling information according to the srvice instance event message to the target service example, and the controlled target srvice instance is finished corresponding operating.

5, method according to claim 1, it is characterized in that: for distributed service logic running environment structure, described step B further comprises: after event scheduler is received event message, judge whether the corresponding business example is positioned on the service logic running environment node at its place, if obtain the corresponding business example in the type of service pond on this service logic running environment node; Otherwise, the service logic running environment node at this event scheduler place is according to the target service logic operation environment node of initial configuration ff to corresponding target service example place, and the message-oriented middleware address that obtains object event scheduler on the target service logic operation environment node, this event scheduler is to the message of object event scheduler transmission message-oriented middleware form, and the object event scheduler obtains the corresponding business example.

6, method according to claim 2 is characterized in that, for distributed service logic running environment structure, described step B further comprises: the srvice instance place logic operation environment node that is obtained is changed to the main business example with this srvice instance.

7, method according to claim 6 is characterized in that, for distributed service logic running environment structure, the srvice instance that is positioned at same logic operation environment node with the main business example is set to the local service example;

This method further comprises following steps:

D, local service example call data access interface;

8, method according to claim 7 is characterized in that,

9, method according to claim 6 is characterized in that, for distributed service logic running environment structure, the srvice instance that is positioned at Different Logic running environment node with the main business example is set to the teleaction service example;

This method further may further comprise the steps:

D1, teleaction service example call data access interface;

10, method according to claim 9 is characterized in that, generates described local data access tool and remote data access instrument in initialization procedure.

11, method according to claim 2, it is characterized in that, for distributed service logic running environment structure, described steps A further comprises: the event scheduler of network protocol stack on any service logic running environment node sends service conversation primary event message.