Many companies are re-engineering their enterprise computing systems to be more effective and productive. However, even these companies must continue to integrate with legacy computing systems at their partners. Consequently, enterprise computing systems must be able to run in a distributed and heterogeneous environment, performing complex single tasks in parallel. This need is increasingly being met through the use of workflow-based applications, i.e. software applications executing specific and defined business processes.
Companies need to be continually more flexible to react to ever-changing business conditions. For example, companies using business process-based workflow applications must have the ability to adapt quickly to changes and/or upgrades of existing business processes. In another example, the time required for execution of business processes must be minimized, and their execution made more resource-efficient. Accordingly, in defining business processes all unnecessary tasks must be eliminated, and all remaining tasks must be performed with the highest degree of parallelism possible.
The drive for efficiency can make business process management inflexible and not configurable to dynamic company-specific needs. For instance, a business process can be defined and represented by a process graph in a workflow builder tool, and then delivered to a customer for storage and execution. Once set, however, the business process graph could not be changed without changing the underlying process definition.
The workflow builder tool conventionally employs block-oriented modeling to avoid error-prone “spaghetti code” and deadlocks found ins so-called free modeling. Block-oriented modeling guides the user and makes sure that only valid processes can be created. Additionally, block-oriented models do not need to be validated (i.e. simulated) in a complex manner, such as when the user spawns several parallel activities and forgets to join them together. Block-oriented modeling uses so-called subprocesses to avoid having to replicate process parts over and over, a technique known in structured programming languages as “re-use.” Also, block-oriented modeling corresponds with standards such as BPML or BPEL4WS, and XML formatting requires the structure of block-oriented modeling.
Block-oriented modeling has several problems. Users may feel restricted in their use of block-oriented modeled business processes, particularly to address local events that occur during runtime. Further, in order to make modeling processes easier and more intuitive, the block structure may also lead to too many copies of reusable process parts within the process definition.
This document describes a system and method for integrating local changes to business processes. The business processes define communication between two or more business applications. In one aspect, a system includes a configuration module that generates a copy of an original business process, and a graphical tool. The graphical tool is configured for generating a graphical representation of the original business process in a graphical user interface, the graphical tool further configured to receive inputs representing changes to the original business process to generate a delta record. The system further includes a blending tool to merge the delta record with the copy of the original business process.
In another aspect, a system for executing business process logic between two more business applications includes a global workflow container storing one or more business processes, each business process having process logic that defines communication between the two or more business applications. The system further includes a local workflow builder for defining a local workflow, the local workflow having one or more local events that include logic for processing information external to the process logic of a business process. The system further includes a local workflow container storing the local workflow, and being accessible by the business process to receive and/or send data from and/or to the local workflow.
In yet another aspect, a method of executing business process logic between two more business applications includes retrieving a business process from a global process container, the business process including global process logic governing communication between the two or more business applications. The method further includes executing a local workflow based on one or more triggers, the local workflow being defined independently from the business process and configured to provide data to at least one step of the business process.
The details of one or more embodiments are set forth in the accompanying drawings and the description below. Other features and advantages will be apparent from the description and drawings, and from the claims.
These and other aspects will now be described in detail with reference to the following drawings.
Like reference symbols in the various drawings indicate like elements.
The systems and techniques described here relate to management of business processes that define a message communication protocol between applications in a heterogeneous system landscape. The business process management system and method is optimally implemented in an exchange infrastructure configured to integrate and drive collaboration between various applications in the landscape using open standards and transport protocols such as XML and HTTP.
The XI 110 is a self-contained, modularized exchange platform for driving collaboration among the components 102, 104. The XI 110 includes a central integration repository and directory storing shared collaboration knowledge. The XI 110 supports open standards such as various standard markup languages like the extensible markup language (XML), web service description language (WSDL), and simple object access protocol (SOAP) to provide an abstraction of technical interfaces for the components 102, 104, and for message-based communications across heterogeneous component interfaces. The self-contained, modularized functions of the XI 110 can be provided as one or more Web services based on standard Internet technology, and therefore can be published, discovered, and accessed within a network of components 102, 104 using open standards.
The integration server 206 includes a runtime engine 214 that provides messaging and business process control at runtime for connecting services and managing the process flow of value chains. The runtime engine 214 runs within a business process manager 299. The business process manager 299 governs execution of business processes by the runtime engine 214 at runtime.
The integration server 206 also includes integration services 216 that require an application-specific implementation. Like the integration repository 202 and integration directory 204, the integration server 206 is configured for deployment within any existing system infrastructure. The integration server 206 is preferably a dedicated server that applies the shared collaboration knowledge of the integration directory 204 of the supported system landscape in a runtime collaboration environment. A runtime workbench 208 allows organizations or users to manage the reliable operation of the XI 110.
The XI 110 also includes various adapters 209 that provide connectivity between the integration server 206 and proprietary applications 211, Web-based services 213, and third party applications 215. The XI 110 can also include Web applications server 210 that provides Web-based applications programmed according to standard computing platforms using web-specific programming languages such as Java and ABAP, for instance. The Web applications server 210 also includes an instance of the runtime engine 214 for providing messaging and business process control between Web-based applications such as Java applications 220 and ABAP applications 222, and other components.
New interfaces for software components can be defined using an application component employing a proxy, which allows the interface for the software component to be implemented locally in the XI 110. Proxies make the communication technology stack transparent to applications, and present an application with a programming language-dependent interface. The proxies can be generated by a proxy generator 218 based on information stored on the integration repository 202. The proxy generator 218 uses the interface information described via a standard Web-based language such as WSDL and XSDL to create platform- and programming language-dependent code in the application development system.
The communication logic can be implemented based on the proxy that represents the interface description of the respective development platform, such as Java, ABAP, and .NET for the web-based applications 213. The proxies convert platform-specific data types into XML and provide access to the component-specific local integration engine. On the outbound side, proxies are generated completely. Outbound proxies can be called via a service invocation provided by an application's developer. On the inbound side, only proxy skeletons need to be generated, as implemented by the receiving application.
The business processes 232 can be implemented as extensible compound Web services executed using a business process engine 274. Each business process 232 is modeled centrally in the integration repository 202. A company or user designs each business process 232 according to its business needs, independently of the technical implementation. There may be several categories of business process templates: i.e. generic business processes, industry-specific processes, and company-specific processes, for example. Each process identifies the Web services that are needed and that must be interconnected.
In one specific implementation, business processes 232 can be defined in a configuration layer 290. Further, the configuration layer 290 can be used to dynamically reconfigure business processes being executed at runtime. An extensible import/export framework provides import/export facilities for other standards or new versions of business process models. The business process engine 274 (
Routing objects 234 are predefined criteria to determine potential receivers of messages that must be distributed between components and business partners during collaborative processing. Information about the routing objects is used for receiver determination to avoid having to process a complete message before distribution. Mappings 236 define required transformations between message interfaces 238, message types, or data types in the integration repository 202. These transformations cover structural conversions and value mappings. Structural conversions are used for semantically equivalent types of messages that are syntactically or structurally different, whereas value mapping may be used when an object is identified by different keys in multiple systems. In a specific implementation, a graphical mapping tool is provided to assist in mapping, and transforming data is based on the Extensible Stylesheet Language Transformation (XSLT) or Java code.
The integration repository 202 is the central point of entry for interface development, storage and retrieval, and includes interfaces 238 that describe all message interfaces of all software components in the system landscape. Accordingly, the interfaces 238 can be implemented on any software component using any technology. Message interfaces are made up of message types, which are in turn made up of data types. The data types can be described using XML Schema Definition Language (XSDL). An example of a data type is “address,” which is used in the message type “Create PO” and can be reused for the message type “Create Invoice.” Interfaces 238 can be arranged according to any classification, such as inbound, outbound and abstract, or synchronous and asynchronous.
The components 240 represent component descriptions that include information about application components, as well as information relating to their dependencies on each other. In a specific implementation, the component descriptions are based on the standard Common Information Model (CIM) of the Distributed Management Taskforce. Since the integration repository 202 includes design-time information, only component-type information, independent of actual installation, is stored as components 240 in the system landscape directory 203. The component descriptions can be added using an API or interactively using a graphical user interface.
The integration directory 204 details information from the integration repository 202 that is specific to the configuration of each component as installed in the system. The configuration-specific collaboration descriptions of the integration directory 204 can be generated automatically from content in the integration repository 202 or manually by a user using a graphical user interface. In one implementation, the integration directory 204 is built on a Java platform and its content is represented via XML using open Internet standards. The integration repository 202 can be upgraded without affecting the integration directory 204 or any runtime collaborative processes. The user then decides which changes should be transferred to the integration directory 204, either as predetermined automatic upgrades or manually via graphical tools.
The integration directory 204 includes configuration-specific descriptions of business scenarios 250, business processes 252, context objects 254, and executable mappings 256. The integration directory 204 also includes descriptions of active Web services 258, and active business partners 260. The integration directory 204 uses a description of the active system landscape 262 from the system landscape directory 203. The business scenarios 250 in the integration directory 204 represent the overall view of the interaction among interfaces and mappings 256 in the context of the actual configuration relevant for the specific implementation. The business processes 252 represents an executable description of all active business processes.
The context objects 254 determine the receivers of a message on a business level. In one specific implementation, the content of a message is used as a context object 254. Other parameters may also be used. Relevant input parameters include the sender, the sender message type, the message to identify the receivers, and the receiver message type. The context object 254 can be described declaratively using XML Path Language (Xpath, i.e. by using a graphical tool) or can be coded in Java. The integration engine 214 at runtime accesses information on the context object 254.
The context objects 254 may use logical terms to describe senders and receivers in order to separate them from the physical address provided by the Web services 258 described in the integration directory 204. The physical address can therefore be changed without changing business-oriented content. Mappings 256 in the integration directory 204 represent mappings required in the active system landscape, in contrast to the integration repository mappings 236 that contains all supported mappings. Some new entries however, such as a new sequence of mappings, can be made only in the integration directory 204 to address additional Web services for mapping, for example. The integration engine 214 accesses the integration directory mappings 256 at runtime.
Context objects 254 provide a unique name for accessing semantically identical payload information. For instance, a context object can provide a unique access name for ‘plant’ for invoice and purchase order. The XPath for ‘plant’ in an invoice can be defined as ‘/A/B/C/plant’ and the XPath for ‘plant’ in a purchase order looks like ‘X/Y/Z/work’. The context object 254 ‘plant’ is assigned to the message interface invoice and purchase order where the XPaths as above mentioned are specified. This makes sure that the XPath for plant is not defined at n different places.
Web services 258 describe interfaces implemented within the current active system landscape, as well as active Web services supported by described business partners 260. As such, information describing Web services 258 can be exchanged with Universal Description, Discovery, and Integration (UDDI) compatible directories or added manually. Each Web service 258 description also provides physical addressing details, access information, and other special attributes such as uniform resource locator (URL), protocol, and security information. In one implementation, the Web services 258 are described in WSDL, and SOAP and ebXML are used as messaging protocols. The integration engine 214 accesses information about the Web services 258 at runtime as well.
The system landscape 262 of the system landscape directory 203 describes the current system landscape that uses the XI 110. The system landscape 262 describes the components that are installed and available on certain machines within the system, the instance or client that was chosen, further information on the installed components, other system landscapes, and so on. The system landscape 262 description is based on an open architecture and can adhere to any widely accepted standard such as the Common Information Model (CIM). Thus, many proprietary and third party components can be configured to automatically register themselves in the system landscape 262 upon being installed within the actual system landscape. Access interfaces to the system landscape 262 description can be based on open standards as well, such as the Web-based Enterprise Management (WBEM) and SOAP standards.
Business partners 262 defines information for business partners of an enterprise, such as names, addresses, and URLs, but may also contain more detailed and sophisticated information. For instance, the business partners 262 may include a description of the message formats that can be directly received and processed, or of security protocols used for safe communications, or trading terms that are employed in the partnership. The kind of information stored in business partners 262 can be governed by enterprise-specific decisions of the enterprise using the XI 110.
The integration directory 204 and the runtime engine 214 form a collaborative runtime environment for executing collaborative business processes. The collaborative runtime environment provides all runtime components relevant for exchanging messages among the connected software components and business partners. The integration server 206 executes the collaborative runtime environment or Web application server 210, either of which can include an instance of the runtime engine 214 in accordance with informational resources provided by the integration directory 204.
The runtime engine 214, which exchanges all messages between the various interconnected components, includes two layers: an integration layer 272 and a messaging and transport layer (MTL) 280. The integration layer 272 includes a business process engine 274 executing centrally modeled business processes, a logical routing service 276 and a mapping service 278. The MTL 280 provides a physical address resolution service 282, a messaging and queuing service 284, a transport service 286 via HTTP, and a database 288. The integration services 216 in the integration server 206 can support the runtime engine 214. An MTL 280 is also included in each instantiation of the runtime engine 214 in Web applications servers 210, as well as in each adapter 209 of the adapter framework connecting to various software components. Each MTL 280 has a role in the execution of the EO protocol, as will be explained further below.
At runtime, business processes 252 are instantiated and executed by the business process engine 274, which executes the respective Web services described in Web services 258 independent of their location according to the business process model. The business process engine 274 is independent of the semantics of the executed business processes 252, and is configured as a mediator and facilitator for business processes 252 to interact with technical components of the runtime system landscape.
A receiving component system 304 may also utilize an inbound proxy 311 rather than an adapter. The configuration and connectivity of the shown receiving component systems 304 is merely exemplary, and it should be noted that such configuration and connectivity could take any number of forms. The pictured example illustrates both asynchronous and synchronous communication. In synchronous communication, routing and physical address resolution is only needed for the request as the response is transferred to the sender, which is already known.
For a given message the logical routing service 276 uses information on the sending application and the message interface to determine receivers and required interfaces by evaluating the corresponding routing rules, as shown at 312. The routing rules are part of the configuration-specific descriptions of the runtime system landscape provided by the integration directory 204, and can be implemented as XPath expressions or Java code. The mapping service 278 determines the required transformations that depend on message, sender, and sender interface, as well as the receiver and receiver interface, at 314. In the case of asynchronous communication, even the message direction is determined to appropriately transform input, output, and fault messages.
After retrieving the required mapping from the integration directory 204, the mapping service 278 can either execute XSLT mappings or Java code (or any combination in a given sequence) to the content of the sent message. Below the integration layer, messaging, queuing, and transport services 284 move the message to the intended or required receiver(s). After the message is transformed into the format expected by each receiver, the physical address of the required receiver service and other relevant attributes are retrieved from the integration directory 204 and mapped to the message, at 316.
A queuing engine (not shown) in the messaging and queuing service 284 stores ingoing, outgoing, erroneous, and work-in-progress messages persistently. The messaging layer of the runtime engine 214 provides queuing functions for the physical decoupling of application components and guarantees messages are delivered exactly once. The transport service 286 enables the runtime engine 214 to act as both a client and server. The transport service 286 implements a client that enables outbound communication and a server that handles inbound communication by accepting incoming documents. Additional server functions can address situations in which the receiver has no server by supporting polling over the transport protocol used. HTTP is preferably used, but other transport protocols may be used as well.
The process definition 506 module utilizes XML objects and correlations to define processes, based on deployment rules imported from XI objects 512 from the integration directory 514. The XI objects 512 are based on the routings and mappings defined for the system runtime configuration 516. The XI objects 512 are also used to define business processes 518 in the integration repository 522, and the design-time configuration 520 of the system landscape.
Business processes 518 are integrated with and linked with other objects and tools in the integration repository 522. Business processes 518, in the form of patterns and templates, can be delivered to customers. Application-specific content can also be delivered. The BPM system 500 includes an import/export framework 526 that imports and exports standards-based adapters for universal connectivity. The BPM system 500 can include an interface for receiving user-specified business process details.
Business process modeling scenarios, called “patterns,” are high-level building blocks that can be combined with each other and with atomic functions such as deadlines, exceptions, etc. of the process engine 504. The following are example patterns:
1) Send and Receive: Sending messages controlled by the process engine 504 is often combined with receive steps that wait for a correlated response message. A receive step should wait for the messages starting with the activation of the associated correlation as a queuing mechanism.
2) Serialization: This pattern can include the following steps: 1. Receive messages and store them locally in the process data context; 2. Keep the data context and start sending received messages when a certain condition has been fulfilled; and 3. Send received messages in a given order respecting dependencies of receivers. This third step can be: a. Without caring about responses/acknowledgements (“fire and forget”); or b. Receiving a response or an acknowledgement (enables serialization). The process engine 504 can be configured to wait for a technical ACK of or business response from a previously-sent message before sending a next message.
3) Transformations/Merge/Split: The process engine 504 transforms messages within the process context. The following transformations can be performed: 1. (N:1) Transform several collected messages to one new message (e.g. transform several invoices to one combined invoice or transform PO header and several PO positions into one PO); 2. (1:N) Transform one message into several other messages (e.g. transform a combined in-voice to invoice respecting the original POs); and 3. (1:1) is a special case of the transformations described above. N:M mappings are also possible if needed.
4) Multicast: The process engine 504 can be configured to calculate the receivers of a message (also using content-based conditions) and to send the message to these receivers, either without regard to responses/acknowledgements (“fire and forget”) or based on receiving a number of responses/acknowledgements. Messages may be sent out in parallel or sequentially.
5) Collect: This pattern uses receive steps in which an arbitrary number of messages can be received. From a process point of view, the end of the collecting scenario can be defined via “push,” (i.e. a certain condition is reached, such as N messages have arrived, a certain deadline has been reached, etc.), or “poll” in which the process engine waits for a special message that indicates the end of collecting.
The process engine 504 uses business processes on the integration server 502. While it is able to communicate with backend processes via messages, the process engine 504 does not interact with the applications, organizational and user management functions in the backend system(s). The process engine 504 uses the messaging layer application, while business workflow uses the application, user, and organizational management of the respective application system. The process engine 504 supports the communication via synchronous outbound interfaces.
Processes will have representations both in the integration repository 522 and the integration directory 514. Process definitions are stored in the integration repository 522. This allows the transport of process definitions to the client systems 503. Processes stored in the integration directory 514 point to an associated process definition in the integration repository 522. Business processes 518 include public parts, such as previously-used interfaces, and private parts, which include the process graph using step types and correlations. Process instances can be stopped and restarted in runtime 508. Process instances can also be restarted from any particular step (e.g. if an error occurs during a certain step, restart from that step).
Each business process, as an XI object that is visible in a navigation tree and usable in links from and to other XI objects, will provide the ability to integrate the process engine 502 in the XI environment. Business processes 518 can use established XI object types, and will not create redundant object types.
A business process 518 “owns” a process interface 708 which reflects all inbound and outbound communication. Interfaces used in the process interface 708 include two types: process-specific interfaces (a special normalizing interface for the process); and mirrored outbound/inbound interfaces of sending/receiving business systems (to avoid the creation of unnecessary interfaces). Mirroring must be done creating a new abstract interface 702 pointing to the same message type as the original interface. Abstract interfaces 702 can be used in an inbound as well as in an outbound role.
Should the process need inbound and outbound messages, a transformation from inbound to outbound can be executed. In addition, process-specific interfaces do not need to have proxies in the attached business systems. This leads to the so-called abstract interfaces 702, which are the only type of interfaces that can be used by the business processes 518. Local interfaces may reference other interfaces 708 (to handle the mirroring) and they also may reference message types 710 (to realize process-specific interfaces). Context objects 704 can be used to access payload information via name or other message content. Data may not be written to context objects 704. Interface mappings 706 are addressed by a business process 518 within the transformation step.
A business scenario 720 may reference one or more business processes 518. One business process occupies one “swim lane” or process flow. Each process is treated as a business system. Actions within process swim lanes are not stored as separate actions that are reusable. An action represents an interface used by a business process 518 as outbound or inbound interface (or both). In a ‘normal’ scenario case, not all interfaces of an action must be a target or source of a connection. In a business process definition, each action represents one interface with inbound and/or outbound semantics and must be used as a target and/or source in a connection.
A graphical tool 808 generates a graphical representation of the original business process 804, for display in a graphical user interface 810. The graphical tool 808 also receives user input commands and instructions from a user input device 812, and represented in the graphical user interface 810 by menu options provided, for example, by a pull-down menu 811. The user input device 812 can be a keyboard connected with a computer that hosts the graphical tool 808. The user input device 812 can also be a mouse or other type of electronic input device. The user input commands and instructions includes instructions for changing the original business process 804. The changes to the original business process 804 are displayed in the graphical user interface 810, as a graphical representation of the changes.
The system 800 further includes a blending tool 814 that receives the changes as a delta record 805 and blends them with the copy of the original business process 804, to create a file of the original plus the delta record 805. The original plus delta business process is then stored in the integration directory, where the original business process 804 is preserved.
In an embodiment, the user can select menu choices provided in the graphical user interface. The menu choices can include process steps that can be inserted into the graphical representation of the copy of the business process. Or, a user may deselect and remove other process steps from the copy of the business process. At 824, user inputs are received. The user inputs represent the desired changes to the business process, and are configured to change the graphical representation of the copy of the business process within the graphical user interface. At 826, a delta record is generated, reflecting the local changes made to the copy of the business process.
At 828, a determination is made whether there are any more changes to be made to the copy of the original business process. This determination can be provided by a particular user input, such as a selection of a “save” command, or other user action. If more changes are to made, the method returns to 824 with the receipt of further user input. Once all changes have been made, at 830 the copy of the original business process is merged with the delta record of the changes. At 832, both the delta record and the copy of the business process are stored in a local repository.
In another embodiment, local events can be defined and used for sending and receiving signals within a process instance of a predefined business process. As shown in
The local workflows 910 are smaller or larger pieces of process logic that are defined independently from the main process logic of a predefined business process 902. Local workflows 910 may have their own address space and their own local container elements, yet also have access to the global process container which contains the predefined, delivered business processes. Local events 908 can also be assigned a local container.
Producers and consumers of local events, i.e. a local application and a global business process, etc., can hand over data to the local event or get data from the local event. Local workflows 910 can be called in parallel and do not affect the structure of the global process logic of a predefined business process 902, although each local workflow can change container elements of the global workflow represented by the process graph 904, and which could influence the output of the global process logic of the predefined business process 902. A local workflow definition may be triggered and instantiated n times, while each local workflow instance has a separate address space within its local container.
A local workflow builder 928 is a local application running on a local host computer that preferably includes graphical interface capabilities. The local workflow builder 928 can be used by a user to generate one or more local workflows. Each local workflow includes one or more local events that are triggered by steps defined in the business process logic in the global process container, or by external events that occur outside the instance of a predefined global business process. Local workflows and local events generated by the local workflow builder 928 are stored in a local workflow container 930. The local workflow container 930 is connected to be accessed from, and have access to, the global process container 926, such that local events and local workflows that are triggered can be executed seamlessly, without changing the basic structure of the global business process logic. The local workflow container 930 can be stored in the same or a different repository than the global process container 926.
A method of using local workflows and local events in a business process management system is shown in
At 936, a trigger may be employed to execute a local workflow and/or local event. The trigger can be defined within a predefined business process, or started by an event external to the process logic of the predefined business process. If the trigger is employed, n local workflows can be instantiated and executed in parallel at 938, each with its own address space in the local workflow container. Accordingly, the use of local events and local workflows allows flexibility in business process management, and yet allow reuse of block-oriented modeling of process parts. Further, local events and local workflows allow for reaction to unexpected external events that are outside the instance of a global process logic. Furthermore, multiple unexpected external events can be handled at the same time.
Although a few embodiments have been described in detail above, other modifications are possible. The logic flows depicted in