WO2012116513A1 - 并发优化的bpmn组合服务执行引擎及方法 - Google Patents

并发优化的bpmn组合服务执行引擎及方法 Download PDF

Info

Publication number
WO2012116513A1
WO2012116513A1 PCT/CN2011/072683 CN2011072683W WO2012116513A1 WO 2012116513 A1 WO2012116513 A1 WO 2012116513A1 CN 2011072683 W CN2011072683 W CN 2011072683W WO 2012116513 A1 WO2012116513 A1 WO 2012116513A1
Authority
WO
WIPO (PCT)
Prior art keywords
file
task
unit
execution
service
Prior art date
Application number
PCT/CN2011/072683
Other languages
English (en)
French (fr)
Inventor
刘旭东
李建欣
方琨
赵永望
杨帆
Original Assignee
北京航空航天大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京航空航天大学 filed Critical 北京航空航天大学
Publication of WO2012116513A1 publication Critical patent/WO2012116513A1/zh

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management

Definitions

  • the present invention relates to a computer composite service technology, and more particularly to a concurrent optimization BPMN composite service execution engine and method. Background technique
  • Combined service modeling and service execution are two important components of the service portfolio lifecycle.
  • business analysts typically use a variety of process modeling languages to describe business processes and perform composite service modeling.
  • Current common process modeling languages include design-oriented languages (such as BPMN, WS-CDL) and implementation-oriented languages (such as BPEL, XPDL).
  • BPMN Business Process Modeling Notation
  • BPMI Business Process Management Initiative
  • OMG object management organization
  • the business process can be modeled by BPMN and bound into a composite service.
  • the business process generated by BPMN's composite service execution engine-driven modeling is executed in sequence. The essence of the line is to transform the business process into a composite service and drive it to perform as required by the modeler.
  • the common BPMN composite service execution method is to convert BPMN modeling primitives into XML SCHEMA files, obtain the execution order of business processes by parsing XML files and drive the sequential execution of their combined services, and provide monitoring ports during execution. Continuously monitor the execution of business processes.
  • the composite service execution engine is mostly in the form of a native program or plug-in, which may result in its deployment depending on a certain platform and the deployment implementation is relatively complex.
  • the composite service execution engine includes multiple loosely coupled modules, such as monitoring modules, it is necessary to consider whether multiple loosely coupled modules can be adaptively deployed across environments.
  • the present invention provides a concurrently optimized BPMN composite service execution engine and method to improve the flexibility and concurrency of BPMN composite service execution engine deployment.
  • the present invention provides a concurrently optimized BPMN composite service execution engine, including:
  • a service interface module configured to obtain a service process description file and input parameter information according to the received message file, and send the service process description file and the input parameter information;
  • a task management module configured to receive the service process description file and the input parameter information sent by the service interface module, perform fragmentation on the service process description file, generate a process fragment file, and slice according to the process a file and the input parameter information to generate a task file, and according to The process fragment file sends the task file;
  • the parsing execution module includes a task receiving unit, a process parsing unit, an instance executing unit, a unit parsing unit, and an adaptive unit;
  • the task receiving unit is configured to receive the task file sent by the task management module, and the process analyzing unit is configured to perform, by using the process fragment file in the task file received by the task receiving unit Parsing, obtaining the execution order, and generating an object model;
  • the instance execution unit is configured to: according to the execution order, according to the input parameter information in the task file and the parsing execution method stored in the primitive parsing unit, invoke a thread to execute the generated by the process parsing unit The object model, and generating an execution result;
  • the adaptive unit is configured to determine, when the number of the task files in the task receiving unit is less than a preset value, invoke a monitoring module by using a web service calling manner, and send the instance execution unit to the monitoring module.
  • the execution result is generated; otherwise, the monitoring module is invoked by using a local process asynchronous calling manner, and the execution result is sent to the monitoring module;
  • the monitoring module includes a runtime monitoring unit, and the runtime monitoring unit is configured to access the database according to the execution result.
  • the task receiving unit includes a task pool, and the task pool is configured to sort a plurality of the task files, and provide the task file to the process parsing unit according to a preset sequence.
  • parsing execution module includes:
  • a thread pool configured to allocate an idle thread to the instance execution unit.
  • monitoring module further includes:
  • the concurrently optimized BPMN combined service execution engine wherein the monitoring module further comprises:
  • the service interface module includes a service interface, a message conversion unit, and a request unit;
  • the message conversion unit is configured to convert the message received by the service interface to obtain the service flow description file and the input parameter information
  • the requesting unit is configured to send the service flow description file and the input parameter information.
  • a receiving unit configured to receive the service flow description file and the input parameter information sent by the service interface module
  • a fragmentation unit configured to slice the business process description file received by the receiving unit, and generate a process fragment file
  • a task file generating unit configured to generate a task file according to the process fragment file and the input parameter information
  • a process scheduling unit configured to perform process scheduling according to the process fragment file, and send the task file.
  • the present invention provides a concurrently optimized BPMN combined service execution method, including: obtaining a business process description file and input parameter information according to the received message file; and segmenting the business process description file to generate a process fragment file Generating a task file according to the process fragment file and the input parameter information, and sending the task file according to the process fragment file;
  • the calling thread executes the object model according to the input parameter information in the task file and the parsing execution method stored in the primitive parsing unit, and generates an execution result;
  • the monitoring service When it is determined that the number of the received task files is less than a preset value, the monitoring service is invoked by using a web service calling manner, and the execution result is sent; otherwise, the local process is asynchronously called. Invoking the monitoring service, and sending the execution result;
  • the database is accessed.
  • the concurrently optimized BPMN combined service execution method wherein the process fragment file in the received task file is parsed, specifically:
  • An idle thread is requested from a thread pool that allocates a thread according to a thread allocation condition to execute the object model.
  • a database link is requested to the database connection pool, and the database is accessed according to the database link assigned by the database connection pool.
  • the concurrently optimized BPMN combined service execution engine and method provided by the present invention can implement parsing and execution of the BPMN combined service, and can perform loosely coupled monitoring according to different service requirements when performing monitoring service call.
  • the module is deployed as multiple services, and can also be bound together to become a service.
  • the service internally uses asynchronous calls to ensure operation, which improves the flexibility of the BPMN composite service execution engine deployment.
  • the concurrent ability of the BPMN composite service execution engine deployment is improved by adopting the asynchronous call mode.
  • FIG. 1 is a schematic structural diagram of a concurrently optimized BPMN combined service execution engine according to Embodiment 1 of the present invention
  • FIG. 2 is a schematic structural diagram of a concurrently optimized BPMN composite service execution engine according to Embodiment 2 of the present invention
  • FIG. 3 is a flow chart of a concurrently optimized BPMN composite service execution method according to Embodiment 3 of the present invention. detailed description
  • the concurrently optimized BPMN combined service execution engine includes a service interface module 11, a task management module 12, and an analysis execution module 13. And monitoring module 14.
  • the service interface module 11 is configured to obtain a business process description file and input parameter information according to the received message file, and send the business process description file and the input parameter information.
  • the service interface module 11 first receives a message file, and the message file may be an external service request message, and the service request message may be a Simple Object Access Protocol (SOAP) message.
  • SOAP Simple Object Access Protocol
  • the service process description file may specifically be an XLM description file in a first-level and second-level format, and the input parameter information is specifically a parameter parameter that is input when the user performs a Web service call.
  • the task management module 12 is configured to receive the service flow description file and the input parameter information sent by the service interface module 11, segment the business process description file, generate a process fragment file, and generate a task file according to the process fragment file and the input parameter information.
  • the task file is sent according to the process shard file. Specifically, the task management module 12 performs process fragmentation according to a pool in the BPMN specification to generate a process score.
  • the file file is implemented by a business entity to implement a process fragment file.
  • the parsing execution module 13 includes a task receiving unit 131, a flow parsing unit 132, an instance executing unit 133, a primitive parsing unit 134, and an adaptive unit 135.
  • the task receiving unit 131 is configured to receive the task file sent by the task management module 12.
  • the flow analyzing unit 132 is configured to parse the process fragment file in the task file received by the task receiving unit 131, obtain an execution order, and generate an object model.
  • the object model may be a java object model, and the process fragment file includes multiple primitives, and the primitives in the process fragment file are mapped to the java object model, and the obtained java object model is also multiple.
  • the instance execution unit 133 is configured to invoke the thread to execute the object model generated by the flow parsing unit 132 according to the input parameter information in the task file and the parsing execution method stored in the primitive parsing unit 134 in accordance with the execution order, and generate an execution result.
  • the object model corresponding to a primitive needs to be executed by one thread, and can be requested by the sending thread to perform a thread call.
  • the adaptive unit 135 is configured to: when the number of the task files in the learned task receiving unit 131 is less than the preset value, invoke the monitoring module 14 by using the web service calling mode, and send the execution result generated by the instance executing unit 133 to the monitoring module 14;
  • the monitoring module 14 is called by the local process asynchronous call mode, and the execution result is sent to the monitoring module 14.
  • the monitoring module 14 includes a runtime monitoring unit 141 for accessing the database 15 based on the results of the execution. Specifically, the database 15 may be queried or stored according to the execution result to monitor the running status of the business process and the result information.
  • the concurrently optimized BPMN composite service execution engine is deployed in the web service container in the form of a web service, and the remote call can be easily implemented without local deployment.
  • the working process can be specifically as follows:
  • the user passes the Web monthly service description language ( The Web Services Description Language (referred to as WSDL) document finds the service invocation mode of the BPMN composite service execution engine.
  • WSDL Web Services Description Language
  • the service interface module 11 of the concurrently optimized BPMN composite service execution engine provides the service description file for the user.
  • the user constructs a suitable SOAP message according to the WSDL document description and sends it to the service interface module 11.
  • the service interface module 11 receives the SOAP message and passes it through the message. Change, convert to the parameter information required by the running layer, that is, obtain the business process description file and input parameter information.
  • the service interface module 11 may specifically send a process execution request to the task management module 12, and after receiving the process execution request, the task management module 12 performs process fragmentation on the pool in the BPMN specification of the process description file obtained by the service interface module 11, and generates a process score.
  • the file file is generated according to the process fragment file and the input parameter information, and the process fragment file and the input parameter information may be encapsulated to form a task file, and the task file may be a task memory object.
  • the entry of the process execution is found according to the process fragment file, and the task file is submitted to the parsing execution module 13 for parsing execution.
  • the task receiving unit 131 of the parsing execution module 13 receives the task file.
  • the task management module 12 may send a service flow execution request to the task receiving unit 131, and the task receiving unit 131 provides the received task file to the process analyzing unit 132.
  • the process parsing unit 132 parses the process fragment file in the task file, maps the primitives in the process fragment file into a java object model, and acquires an execution order.
  • the instance execution unit 133 requests the system thread to execute the java object model according to the execution order, and obtains the behavior semantics and attributes of the java object model according to the parsing execution method stored in the primitive parsing unit 134, and executes according to this, and executes execution. result.
  • the execution result of the java object model is transmitted to the monitoring module 14 by the execution of the java object model.
  • the adaptive unit 135 can adaptively select the monitoring module 14 according to the number of task files in the task receiving unit 131. Call method.
  • the adaptive unit 135 determines that the number of the task files in the learned task receiving unit 131 is less than the preset value, that is, when the execution request of the business process is small, the monitoring module 14 is invoked by using the web service calling mode, and the instance execution is sent to the monitoring module 14.
  • the execution result generated by the unit 133 otherwise, the business process needs to provide high concurrency capability, the local module asynchronous call mode is used to invoke the monitoring module 14, and the execution result is sent to the monitoring module 14.
  • the adaptation unit 135 can also be manually configured through a configuration file. When the number of business process execution requests is large, the calling mode can be manually controlled in the configuration item of the configuration file to improve the concurrent capability.
  • the monitoring module 14 accesses the database 15 according to the execution result, and performs data query or storage.
  • the execution of the BPMN composite service can be based on different industries It is required to deploy the ⁇ coupled monitoring module 14 as multiple services, and can also be bound together to become a service.
  • the service internally uses asynchronous calling to ensure operation, and based on such environment, software is released, discovered, and bound. Flexible and flexible, it can meet the large user service request volume, and improve the flexibility of BPMN composite service execution engine deployment. Moreover, when the business demand is large, the concurrent ability of the BPMN composite service execution engine deployment is improved by adopting the asynchronous call mode.
  • FIG. 2 is a schematic structural diagram of a concurrently optimized BPMN combined service execution engine according to Embodiment 2 of the present invention.
  • the task receiving unit 131 may specifically include a task pool 136, where the task pool 136 is configured to The plurality of task files are sorted, and the task file is provided to the flow analyzing unit 132 in accordance with a preset order.
  • the task file is managed by the task pool 136.
  • the task file in the task pool 136 can be stored in the form of a queue.
  • the task file can be executed according to the first input first output (FIFO) execution mode. It is supplied to the flow analysis unit 132 for analysis.
  • FIFO first input first output
  • the execution order of the task file in the task pool 136 can be set according to the actual business process requirements, not limited to this embodiment. .
  • the number of the business process execution requests executed at the same time can be controlled, and the task is parsed in the task pool 136 by the task, and is provided to the process parsing unit 132 in a certain order. Ensure that the BPMN composite service execution engine does not get stuck in the private state due to the excessive number of business process execution requests, and improves the concurrency of the BPMN composite service execution engine.
  • the parsing execution module 13 may specifically include a thread pool 137 for allocating idle threads to the instance execution unit 133.
  • the thread pool 137 manages the threads.
  • the instance execution unit 133 needs to call the thread to execute the object model, it is not necessary to directly send a thread request to the system, but the thread execution is performed on the instance execution unit 133 through the thread pool 137.
  • the idle thread is provided to the instance execution unit 133, and the thread released by the instance execution unit 133 is returned to the thread pool 137, avoiding a large number of repeated on-site applications and release, further improving BPMN The concurrency of the composite service execution engine.
  • the monitoring module 14 may further include a database connection pool 142 for storing links of the database 15 and allocating links.
  • the corresponding link of the database 15 is managed by the database connection pool 142.
  • the link is obtained from the database connection pool 142 uniformly.
  • the link is not directly released, but the link is returned to the database connection pool. 142, to avoid a sharp drop in performance caused by frequent access to the database 15, further improving the concurrency of the BPMN composite service execution engine.
  • the service interface module 11 may specifically include a service interface, a message conversion unit 111, and a request unit 112.
  • the message conversion unit 111 is configured to convert the message received by the service interface to obtain a service process description file and input parameter information, where the service interface is specifically a Web service interface.
  • the request unit 112 is configured to send the business process description file and the input parameter information.
  • the task management module 12 may specifically include a receiving unit 121, a fragmentation unit 122, a task file generating unit 123, and a flow scheduling unit 124.
  • the receiving unit 121 is configured to receive the service flow description file and the input parameter information sent by the service interface module 11, and the fragmentation unit 122 is configured to slice the service flow description file received by the receiving unit 121, generate a process fragment file, and a task file.
  • the generating unit 123 is configured to generate a task file according to the process fragment file and the input parameter information, and the process scheduling unit 124 is configured to perform process scheduling according to the process fragment file, and send the task file.
  • FIG. 3 is a flowchart of a concurrently optimized BPMN combined service execution method according to Embodiment 3 of the present invention, where the concurrent optimized BPMN combined service execution method includes:
  • Step 10 Obtain a business process description file and input parameter information according to the received message file.
  • Step 20 Fragment the service process description file, generate a process fragment file, generate a task file according to the process fragment file and the input parameter information, and send the task file according to the process fragment file;
  • Step 30 Parse the process fragment file in the received task file, obtain an execution sequence, and generate an object model.
  • Step 40 In accordance with the execution order, according to the input parameter information in the task file and the parsing execution method stored in the primitive parsing unit, the calling thread executes the object model, and generates an execution result;
  • Step 50 Determine whether the number of the task files in the task receiving unit is less than a preset value; if it is determined that the number of the received task files is less than the preset value, perform step 60, invoke the monitoring service by using a web service calling manner, and send Results of the;
  • step 70 is executed, and the monitoring service is invoked by using a local process asynchronous call mode, and the execution result is sent;
  • Step 80 Access the database according to the execution result.
  • the concurrently optimized BPMN combined service execution method provided in this embodiment can implement the execution of the BPMN composite service, and when the monitoring service is invoked, the loosely coupled monitoring module can be deployed as multiple services according to different service requirements. It can also be bundled together to become a service.
  • the service internally uses asynchronous calls to ensure operation, which improves the flexibility of BPMN composite service execution engine deployment. Moreover, when the business demand is large, the concurrent ability of the BPMN composite service execution engine deployment is improved by adopting the asynchronous call mode.
  • parsing the process fragment file in the received task file may be: parsing the process fragment file in the received task file according to a preset order.
  • the task file may be managed by the setting of the task pool, and the received task file is stored in the task pool, and the task file is parsed according to a certain preset order, which may be in accordance with the execution order of the FIFO, or may be The manner in which the priority is set is not limited to this embodiment.
  • the calling thread execution object model may specifically be: requesting an idle thread from the thread pool, and the thread pool allocates a thread according to the thread allocation status to execute the object model.
  • accessing the database specifically includes: requesting a database from the database connection pool Link, accessing the database based on the database link assigned by the database connection pool.
  • the database link is managed uniformly by the data connection pool, and by requesting the database link to the data connection pool, the system performance degradation due to excessive requests is avoided.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Exchange Systems With Centralized Control (AREA)

Description

并发优化的 BPMN组合服务执行引擎及方法
技术领域
本发明涉及计算机组合服务技术, 尤其涉及一种并发优化的 BPMN组合 服务执行引擎及方法。 背景技术
随着 Internet和网络技术的快速发展, 面向服务体系结构 SOA等网络化 服务软件模式得到了广泛应用, 特别是 Web服务作为 SOA的最佳实践, 极 大的推动了 SOA在金融、 电信、 电子商务等领域的发展。 在实际的业务应用 中独立的 Web服务功能有限, 不可能完成复杂的业务需求, 一些多方参与的 复杂业务流程需要集成已存在的 Web服务形成更高级服务组合来完成。 Web 服务组合作为实现灵活的跨组织应用集成和资源共享的核心技术得到了许多 研究机构的共同关注, 成为新兴的研究热点。
组合服务建模和服务的执行, 是服务组合生命周期中的两个重要组成部 分。 在业务建模阶段, 业务分析人员通常使用各种流程建模语言描述业务流 程, 进行组合服务建模。 目前常见的流程建模语言包括面向设计的语言 (如 BPMN, WS-CDL )和面向执行的语言 (如 BPEL、 XPDL )等。 业务流程建 模标记( Business Process Modeling Notation, 简称 BPMN )是由业务流程管理 计划组织 ( Business Process Management Initiative, 简称 BPMI )提出的用类 似流程图的形式描述业务流程的标准, 目前由对象管理组织 ( Object Management Group , 简称 OMG ) 来维护管理。
可以通过 BPMN的建模方式对商业业务流程进行建模并且将其绑定成为 组合服务, 通过 BPMN的组合服务执行引擎驱动建模产生的业务流程按序执 行,其实质为将业务流程转换为组合服务并驱动其按照建模人员的要求执行。 常见的 BPMN组合服务执行方法是将 BPMN建模图元转换为 XML SCHEMA 文件, 通过解析 XML文件获得业务流程的执行顺序并且驱动其组合服务的 按序执行, 同时在执行过程中需要提供监控端口来不间断地监控业务流程的 执行状况。
在 SOA体系中, 由于网络环境是松散耦合的, 同时也是跨组织跨平台应 一、 引擎部署不够灵活
组合服务执行引擎大多为本地程序或者插件形式, 这样就导致其部署可 能会依赖于某种平台并且部署实施相对比较复杂。 当组合服务执行引擎包括 多个松耦合模块, 例如监控模块, 需要考虑多个松耦合模块能否进行自适应 的跨环境部署。
二、 并发能力有限
没有考虑并发能力的因素, 需要借鉴传统软件应用中的一些并发解决方 法来提高系统的并发能力, 同时 BPMN组合服务执行方法可能涉及到不同服 务交互的问题, 这样的同步调用也有可能导致并发能力的下降。 发明内容
本发明提供一种并发优化的 BPMN组合服务执行引擎及方法, 以提高 BPMN组合服务执行引擎部署的灵活性和并发能力。
本发明提供一种并发优化的 BPMN组合服务执行引擎, 包括:
服务接口模块, 用于根据接收到的报文文件, 获得业务流程描述文件和 输入参数信息, 并将所述业务流程描述文件和所述输入参数信息发送;
任务管理模块, 用于接收所述服务接口模块发送的所述业务流程描述文 件和所述输入参数信息, 对所述业务流程描述文件进行分片, 生成流程分片 文件, 根据所述流程分片文件和所述输入参数信息生成任务文件, 并根据所 述流程分片文件将所述任务文件发送;
解析执行模块, 包括任务接收单元、 流程解析单元、 实例执行单元、 图 元解析单元和自适应单元;
所述任务接收单元, 用于接收所述任务管理模块发送的所述任务文件; 所述流程解析单元, 用于对所述任务接收单元接收的所述任务文件中的 所述流程分片文件进行解析, 获得执行顺序, 并生成对象模型;
所述实例执行单元, 用于按照所述执行顺序, 根据所述任务文件中的所 述输入参数信息和所述图元解析单元中存储的解析执行方法, 调用线程执行 所述流程解析单元生成的所述对象模型, 并产生执行结果;
所述自适应单元, 用于判断获知所述任务接收单元中的所述任务文件的 数量小于预设值时, 采用 Web服务调用方式调用监控模块, 并向所述监控模 块发送所述实例执行单元产生的所述执行结果; 否则, 采用本地进程异步调 用方式调用所述监控模块, 并向所述监控模块发送所述执行结果;
监控模块, 包括运行时监控单元, 所述运行时监控单元用于根据所述执 行结果, 对数据库进行访问。
如上所述的并发优化的 BPMN组合服务执行引擎, 其中,
所述任务接收单元包括任务池, 所述任务池用于将多个所述任务文件进 行排序, 并根据预设顺序向所述流程解析单元提供所述任务文件。
如上所述的并发优化的 BPMN组合服务执行引擎, 其中, 所述解析执行 模块包括:
线程池, 用于为所述实例执行单元分配空闲线程。
如上所述的并发优化的 BPMN组合服务执行引擎, 其中, 所述监控模块 还包括:
数据库连接池, 用于存储所述数据库的链接, 并对所述链接进行分配。 如上所述的并发优化的 BPMN组合服务执行引擎, 其中, 所述监控模块 还包括: 所述服务接口模块包括服务接口、 报文转换单元和请求单元;
所述报文转换单元用于对所述服务接口接收的所述报文进行转换, 以获 得所述业务流程描述文件和所述输入参数信息;
所述请求单元用于将所述业务流程描述文件和所述输入参数信息发送。 如上所述的并发优化的 BPMN组合服务执行引擎, 其中, 所述任务管理 模块包括:
接收单元, 用于接收所述服务接口模块发送的所述业务流程描述文件和 所述输入参数信息;
分片单元, 用于对所述接收单元接收到的所述业务流程描述文件进行分 片, 生成流程分片文件;
任务文件生成单元, 用于根据所述流程分片文件和所述输入参数信息生 成任务文件;
流程调度单元, 用于根据所述流程分片文件进行流程调度, 并将所述任 务文件发送。
本发明提供一种并发优化的 BPMN组合服务执行方法, 包括: 根据接收到的报文文件, 获得业务流程描述文件和输入参数信息; 对所述业务流程描述文件进行分片, 生成流程分片文件, 根据所述流程 分片文件和所述输入参数信息生成任务文件, 并根据所述流程分片文件将所 述任务文件发送;
对接收到的所述任务文件中的所述流程分片文件进行解析, 获得执行顺 序, 并生成对象模型;
按照所述执行顺序, 根据所述任务文件中的所述输入参数信息和所述图 元解析单元中存储的解析执行方法, 调用线程执行所述对象模型, 并产生执 行结果;
判断获知接收到的所述任务文件的数量小于预设值时, 采用 Web服务调 用方式调用监控服务, 并发送所述执行结果; 否则, 采用本地进程异步调用 方式调用所述监控服务, 并发送所述执行结果;
根据所述执行结果, 对数据库进行访问。
如上所述的并发优化的 BPMN组合服务执行方法, 其中, 所述对接收到 的所述任务文件中的所述流程分片文件进行解析, 具体为:
按照预设顺序对接收到的所述任务文件中的所述流程分片文件进行解 如上所述的并发优化的 BPMN组合服务执行方法, 其中, 所述调用线程 执行所述对象模型, 具体为:
向线程池请求空闲线程, 所述线程池根据线程分配状况分配线程, 以执 行所述对象模型。
如上所述的并发优化的 BPMN组合服务执行方法, 其中, 所述对数据库 进行访问, 具体为:
向数据库连接池请求数据库链接, 根据数据库连接池分配的所述数据库 链接对所述数据库进行访问。
由上述技术方案可知, 本发明提供的并发优化的 BPMN组合服务执行引 擎及方法, 可以实现对 BPMN组合服务的解析执行, 且在进行监控服务调用 时, 可以根据不同的业务需求将松耦合的监控模块部署为多个服务, 同时也 能够绑定在一起成为一个服务, 服务内部使用异步调用的方式保证运行, 提 高了 BPMN组合服务执行引擎部署的灵活性。 而且, 当业务需求量大时, 通 过采用异步调用方式, 提高了 BPMN组合服务执行引擎部署的并发能力。 附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案, 下面将对实 施例或现有技术描述中所需要使用的附图作简单地介绍, 显而易见地, 下面 描述中的附图仅仅是本发明的一些实施例, 对于本领域普通技术人员来讲, 在不付出创造性劳动性的前提下, 还可以根据这些附图获得其他的附图。 图 1为发明实施例一提供的并发优化的 BPMN组合服务执行引擎结构示 意图;
图 2为本发明实施例二提供的并发优化的 BPMN组合服务执行引擎结构 示意图;
图 3为本发明实施例三提供的并发优化的 BPMN组合服务执行方法流程 图。 具体实施方式
下面将结合本发明实施例中的附图, 对本发明实施例中的技术方案进行 清楚、 完整地描述, 显然, 所描述的实施例仅仅是本发明一部分实施例, 而 不是全部的实施例。 基于本发明中的实施例, 本领域普通技术人员在没有做 出创造性劳动前提下所获得的所有其他实施例, 都属于本发明保护的范围。
实施例一
图 1为发明实施例一提供的并发优化的 BPMN组合服务执行引擎结构示 意图, 如图 1所示, 该并发优化的 BPMN组合服务执行引擎包括服务接口模 块 11、 任务管理模块 12、 解析执行模块 13和监控模块 14。 服务接口模块 11 用于根据接收到的报文文件, 获得业务流程描述文件和输入参数信息, 并将 业务流程描述文件和输入参数信息发送。 在实际应用中, 服务接口模块 11首 先接收报文文件, 报文文件具体可以为外部的服务请求消息, 该服务请求消 息可以为简单对象访问协议 ( Simple Object Access Protocol , 简称 SOAP ) 消 息。 业务流程描述文件具体可以为一二级制形式的 XLM描述文件, 输入参 数信息具体为用户进行 Web服务调用时输入的必要的参数信息。任务管理模 块 12用于接收服务接口模块 11发送的业务流程描述文件和输入参数信息, 对业务流程描述文件进行分片, 生成流程分片文件, 根据流程分片文件和输 入参数信息生成任务文件, 并根据流程分片文件将任务文件发送。 具体的, 任务管理模块 12按照 BPMN规范中的池( pool )进行流程分片, 生成流程分 片文件, 以实现一个流程分片文件被一个业务实体执行, 不同 pool的业务实 体能够同时运行, 根据流程分片文件找到流程执行的入口。 解析执行模块 13 包括任务接收单元 131、 流程解析单元 132、 实例执行单元 133、 图元解析单 元 134和自适应单元 135。 任务接收单元 131用于接收任务管理模块 12发送 的任务文件。 流程解析单元 132用于对任务接收单元 131接收的任务文件中 的流程分片文件进行解析, 获得执行顺序, 并生成对象模型。 具体的, 对象 模型可以为 java对象模型, 流程分片文件中包含多个图元, 将流程分片文件 中的图元映射为 java对象模型, 所获得的 java对象模型也为多个。 解析获取 执行顺序,就可以按照执行顺序对 java对象模型进行执行。实例执行单元 133 用于按照执行顺序, 根据任务文件中的输入参数信息和图元解析单元 134中 存储的解析执行方法, 调用线程执行流程解析单元 132生成的对象模型, 并 产生执行结果。 具体的, 一个图元对应的对象模型需要通过一个线程来执行, 可以向发送线程请求, 以进行线程调用。 自适应单元 135用于判断获知任务 接收单元 131中的任务文件的数量小于预设值时, 采用 web服务调用方式调 用监控模块 14, 并向监控模块 14发送实例执行单元 133产生的执行结果; 否则, 采用本地进程异步调用方式调用监控模块 14, 并向监控模块 14发送 执行结果。 监控模块 14包括运行时监控单元 141 , 运行时监控单元 141用于 根据执行结果, 对数据库 15 进行访问。 具体的可以根据执行结果对数据库 15进行查询或者存储操作, 以监控业务流程的运行情况以及结果信息。
该并发优化的 BPMN组合服务执行引擎以 web服务的形式部署在 web 服务容器中, 可以简易地实现远程调用, 而不需要本地部署, 其工作过程具 体可以为: 用户通过 Web月良务描述语言 ( Web Services Description Language, 简称 WSDL )文档找到 BPMN组合服务执行引擎的服务调用方式, 具体的, 该并发优化的 BPMN组合服务执行引擎的服务接口模块 11为用户提供服务 说明文件。 用户根据 WSDL文档说明, 构造合适的 SOAP消息, 并将其发送 给服务接口模块 11。 服务接口模块 11接收到 SOAP消息, 将其通过报文转 换, 转换成为运行层所需的参数信息, 即获取业务流程描述文件和输入参数 信息。 服务接口模块 11具体可以向任务管理模块 12发送流程执行请求, 任 务管理模块 12接收到流程执行请求后, 对服务接口模块 11获得的流程描述 文件 BPMN规范中的 pool进行流程分片, 生成流程分片文件,根据流程分片 文件和输入参数信息生成任务文件, 具体可以将流程分片文件和输入参数信 息进行封装, 以形成任务文件, 任务文件具体可以为任务内存对象。 根据流 程分片文件找到流程执行的入口,将任务文件交给解析执行模块 13去解析执 行。 解析执行模块 13的任务接收单元 131接收任务文件, 具体的, 任务管理 模块 12可以先向的任务接收单元 131发送业务流程执行请求,任务接收单元 131将接收到的任务文件提供给流程解析单元 132,流程解析单元 132对任务 文件中的流程分片文件进行解析, 将流程分片文件中的图元映射成 java对象 模型, 并获取执行顺序。 实例执行单元 133根据执行顺序, 请求系统线程对 java对象模型进行执行, 具体根据图元解析单元 134中存储的解析执行方法 来获取 java对象模型的行为语义和属性, 据此来执行, 并产生执行结果。 在 解析执行过程中, 每个线程执行完的 java对象模型后都会将执行结果传递给 监控模块 14, 自适应单元 135可以根据任务接收单元 131中的任务文件的数 量, 自适应选择监控模块 14的调用方式。 自适应单元 135判断获知任务接收 单元 131中的任务文件的数量小于预设值时,即业务流程的执行请求较少时, 采用 web服务调用方式调用监控模块 14, 并向监控模块 14发送实例执行单 元 133产生的执行结果; 否则, 业务流程需要提供高并发能力, 采用本地进 程异步调用方式调用监控模块 14, 并向监控模块 14发送执行结果。 自适应 单元 135还可以通过配置文件进行手动配置。 当业务流程执行请求量比较大 时, 可以在配置文件的配置项中手动控制调用方式, 以提高并发能力。 监控 模块 14根据执行结果, 对数据库 15访问, 进行数据查询或存储。
本实施例提供的并发优化的 BPMN 组合服务执行引擎, 可以实现对
BPMN组合服务的执行, 且通过自适应单元 135的设置, 可以根据不同的业 务需求将^ ^耦合的监控模块 14部署为多个服务, 同时也能够绑定在一起成为 一个服务, 服务内部使用异步调用的方式保证运行, 基于这样的环境进行软 件的发布、 发现、 绑定和调用灵活, 能够满足较大的用户业务请求量, 提高 了 BPMN组合服务执行引擎部署的灵活性。 而且, 当业务需求量大时, 通过 采用异步调用方式, 提高了 BPMN组合服务执行引擎部署的并发能力。
实施例二
图 2为本发明实施例二提供的并发优化的 BPMN组合服务执行引擎结构 示意图, 如图 2所示, 在本实施例中, 任务接收单元 131具体可以包括任务 池 136, 任务池 136用于将多个任务文件进行排序, 并根据预设顺序向流程 解析单元 132提供任务文件。通过任务池 136对接收到的任务文件进行管理, 任务池 136中的任务文件可以以队列的形式进行存储, 具体可以依照先入先 出 ( First Input First Output, 简称 FIFO ) 的执行方式, 将任务文件提供给流 程解析单元 132进行解析。 还可以为任务文件设定优先权, 根据优先权为流 程解析单元 132提供任务文件, 任务池 136的中任务文件的执行顺序可以按 照实际的业务流程需要来设定, 不以本实施例为限。
通过任务池 136的设置, 作为业务流程执行请求的緩冲, 可以控制同时 执行的业务流程执行请求数量, 通过先将任务放在任务池 136中, 通过一定 的顺序提供给流程解析单元 132解析, 保证 BPMN组合服务执行引擎不会因 为业务流程执行请求数量过多而陷入阻塞家私状态, 提高了 BPMN组合服务 执行引擎的并发能力。
在本实施例中, 解析执行模块 13具体可以包括线程池 137, 线程池 137 用于为实例执行单元 133分配空闲线程。 线程池 137对线程进行管理, 当实 例执行单元 133需要调用线程来执行对象模型时, 不需要直接向系统发送线 程请求,而通过线程池 137对实例执行单元 133进行线程分配,将线程池 137 中的空闲线程提供给实例执行单元 133 , 并将实例执行单元 133释放的线程 返还线程池 137, 避免了大量重复的现场申请以及释放, 进一步提高 BPMN 组合服务执行引擎的并发能力。
在本实施例中, 监控模块 14还可以包括数据库连接池 142, 数据库连接 池 142用于存储数据库 15 的链接, 并对链接进行分配。 通过数据库连接池 142管理数据库 15的相应链接, 当需要连接至数据库 15时, 统一从数据库 连接池 142中获取链接, 当使用结束后, 并不直接释放链接, 而是将链接返 回给数据库连接池 142, 以避免对数据库 15的频繁访问造成的性能的急剧下 降, 进一步提高 BPMN组合服务执行引擎的并发能力。
在本实施例中, 服务接口模块 11具体可以包括服务接口、报文转换单元 111和请求单元 112。报文转换单元 111用于对服务接口接收的报文进行转换, 以获得业务流程描述文件和输入参数信息, 服务接口具体为 Web服务接口。 请求单元 112用于将业务流程描述文件和输入参数信息发送。
在本实施例中, 任务管理模块 12具体可以包括接收单元 121、 分片单元 122、 任务文件生成单元 123和流程调度单元 124。 接收单元 121用于接收服 务接口模块 11发送的业务流程描述文件和输入参数信息,分片单元 122用于 对接收单元 121接收到的业务流程描述文件进行分片, 生成流程分片文件, 任务文件生成单元 123 用于根据流程分片文件和输入参数信息生成任务文 件, 流程调度单元 124用于根据流程分片文件进行流程调度, 并将任务文件 发送。
实施例三
图 3为本发明实施例三提供的并发优化的 BPMN组合服务执行方法流程 图, 该并发优化的 BPMN组合服务执行方法包括:
步骤 10、 根据接收到的报文文件, 获得业务流程描述文件和输入参数信 自 ·
步骤 20、 对业务流程描述文件进行分片, 生成流程分片文件, 根据流程 分片文件和输入参数信息生成任务文件, 并根据流程分片文件将任务文件发 送; 步骤 30、 对接收到的任务文件中的流程分片文件进行解析, 获得执行顺 序, 并生成对象模型;
步骤 40、 按照执行顺序, 根据任务文件中的输入参数信息和图元解析单 元中存储的解析执行方法, 调用线程执行对象模型, 并产生执行结果;
步骤 50、 判断任务接收单元中的任务文件的数量是否小于预设值; 若判断获知接收到的任务文件的数量小于预设值, 则执行步骤 60、 采用 web服务调用方式调用监控服务, 并发送执行结果;
否则, 执行步骤 70、 采用本地进程异步调用方式调用监控服务, 并发送 执行结果;
步骤 80、 根据执行结果, 对数据库进行访问。
本实施例提供的并发优化的 BPMN 组合服务执行方法, 可以实现对 BPMN组合服务的执行, 且在进行监控服务调用时, 可以根据不同的业务需 求将松耦合的监控模块部署为多个服务, 同时也能够绑定在一起成为一个服 务, 服务内部使用异步调用的方式保证运行, 提高了 BPMN组合服务执行引 擎部署的灵活性。 而且, 当业务需求量大时, 通过采用异步调用方式, 提高 了 BPMN组合服务执行引擎部署的并发能力。
在本实施例中, 对接收到的任务文件中的流程分片文件进行解析具体可 以为: 按照预设顺序对接收到的任务文件中的流程分片文件进行解析。 具体 可以通过任务池的设置对任务文件进行管理, 将接收到的任务文件在任务池 中进行存储, 并按照一定的预设顺序对任务文件进行解析, 具体可以按照 FIFO的执行顺序, 也可以通过设定优先权的方式, 不以本实施例为限。
在本实施例中, 调用线程执行对象模型具体可以为: 向线程池请求空闲 线程, 线程池根据线程分配状况分配线程, 以执行对象模型。 通过向线程池 请求线程来执行对象模型, 避免了对系统大量重复的现场申请及释放, 提高 了系统的并发能力。
在本实施例中, 对数据库进行访问具体为: 向数据库连接池请求数据库 链接, 根据数据库连接池分配的数据库链接对数据库进行访问。 由数据连接 池统一对数据库链接进行管理, 通过向数据连接池请求数据库链接, 避免了 由于请求过多造成的系统性能的下降。
最后应说明的是: 以上实施例仅用以说明本发明的技术方案, 而非对其 限制; 尽管参照前述实施例对本发明进行了详细的说明, 本领域的普通技术 人员应当理解: 其依然可以对前述各实施例所记载的技术方案进行修改, 或 者对其中部分技术特征进行等同替换; 而这些修改或者替换, 并不使相应技 术方案的本质脱离本发明各实施例技术方案的精神和范围。

Claims

权 利 要 求
1、 一种并发优化的 BPMN组合服务执行引擎, 其特征在于, 包括: 服务接口模块, 用于根据接收到的报文文件, 获得业务流程描述文件和 输入参数信息, 并将所述业务流程描述文件和所述输入参数信息发送;
任务管理模块, 用于接收所述服务接口模块发送的所述业务流程描述文 件和所述输入参数信息, 对所述业务流程描述文件进行分片, 生成流程分片 文件, 根据所述流程分片文件和所述输入参数信息生成任务文件, 并根据所 述流程分片文件将所述任务文件发送;
解析执行模块, 包括任务接收单元、 流程解析单元、 实例执行单元、 图 元解析单元和自适应单元;
所述任务接收单元, 用于接收所述任务管理模块发送的所述任务文件; 所述流程解析单元, 用于对所述任务接收单元接收的所述任务文件中的 所述流程分片文件进行解析, 获得执行顺序, 并生成对象模型;
所述实例执行单元, 用于按照所述执行顺序, 根据所述任务文件中的所 述输入参数信息和所述图元解析单元中存储的解析执行方法, 调用线程执行 所述流程解析单元生成的所述对象模型, 并产生执行结果;
所述自适应单元, 用于判断获知所述任务接收单元中的所述任务文件的 数量小于预设值时, 采用 Web服务调用方式调用监控模块, 并向所述监控模 块发送所述实例执行单元产生的所述执行结果; 否则, 采用本地进程异步调 用方式调用所述监控模块, 并向所述监控模块发送所述执行结果;
监控模块, 包括运行时监控单元, 所述运行时监控单元用于根据所述执 行结果, 对数据库进行访问。
2、 根据权利要求 1所述的并发优化的 BPMN组合服务执行引擎, 其特 征在于:
所述任务接收单元包括任务池, 所述任务池用于将多个所述任务文件进 行排序, 并根据预设顺序向所述流程解析单元提供所述任务文件。
3、 根据权利要求 1所述的并发优化的 BPMN组合服务执行引擎, 其特 征在于, 所述解析执行模块包括:
线程池, 用于为所述实例执行单元分配空闲线程。
4、 根据权利要求 1所述的并发优化的 BPMN组合服务执行引擎, 其特 征在于, 所述监控模块还包括:
数据库连接池, 用于存储所述数据库的链接, 并对所述链接进行分配。
5、 根据权利要求 1-4任一所述的并发优化的 BPMN组合服务执行引擎, 其特征在于, 所述监控模块还包括:
所述服务接口模块包括服务接口、 报文转换单元和请求单元;
所述报文转换单元用于对所述服务接口接收的所述报文进行转换, 以获 得所述业务流程描述文件和所述输入参数信息;
所述请求单元用于将所述业务流程描述文件和所述输入参数信息发送。
6、 根据权利要求 5所述的并发优化的 BPMN组合服务执行引擎, 其特 征在于, 所述任务管理模块包括:
接收单元, 用于接收所述服务接口模块发送的所述业务流程描述文件和 所述输入参数信息;
分片单元, 用于对所述接收单元接收到的所述业务流程描述文件进行分 片, 生成流程分片文件;
任务文件生成单元, 用于根据所述流程分片文件和所述输入参数信息生 成任务文件;
流程调度单元, 用于根据所述流程分片文件进行流程调度, 并将所述任 务文件发送。
7、 一种并发优化的 BPMN组合服务执行方法, 其特征在于, 包括: 根据接收到的报文文件, 获得业务流程描述文件和输入参数信息; 对所述业务流程描述文件进行分片, 生成流程分片文件, 根据所述流程 分片文件和所述输入参数信息生成任务文件, 并根据所述流程分片文件将所 述任务文件发送;
对接收到的所述任务文件中的所述流程分片文件进行解析, 获得执行顺 序, 并生成对象模型;
按照所述执行顺序, 根据所述任务文件中的所述输入参数信息和所述图 元解析单元中存储的解析执行方法, 调用线程执行所述对象模型, 并产生执 行结果;
判断获知接收到的所述任务文件的数量小于预设值时, 采用 Web服务调 用方式调用监控服务, 并发送所述执行结果; 否则, 采用本地进程异步调用 方式调用所述监控服务, 并发送所述执行结果;
根据所述执行结果, 对数据库进行访问。
8、 根据权利要求 7所述的并发优化的 BPMN组合服务执行方法, 其特 征在于, 所述对接收到的所述任务文件中的所述流程分片文件进行解析, 具 体为:
按照预设顺序对接收到的所述任务文件中的所述流程分片文件进行解 析。
9、 根据权利要求 7所述的并发优化的 BPMN组合服务执行方法, 其特 征在于, 所述调用线程执行所述对象模型, 具体为:
向线程池请求空闲线程, 所述线程池根据线程分配状况分配线程, 以执 行所述对象模型。 征在于, 所述对数据库进行访问, 具体为:
向数据库连接池请求数据库链接, 根据数据库连接池分配的所述数据库 链接对所述数据库进行访问。
PCT/CN2011/072683 2011-02-28 2011-04-12 并发优化的bpmn组合服务执行引擎及方法 WO2012116513A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201110047642.3 2011-02-28
CN201110047642.3A CN102650953B (zh) 2011-02-28 2011-02-28 并发优化的bpmn组合服务执行引擎及方法

Publications (1)

Publication Number Publication Date
WO2012116513A1 true WO2012116513A1 (zh) 2012-09-07

Family

ID=46692961

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2011/072683 WO2012116513A1 (zh) 2011-02-28 2011-04-12 并发优化的bpmn组合服务执行引擎及方法

Country Status (2)

Country Link
CN (1) CN102650953B (zh)
WO (1) WO2012116513A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104699528A (zh) * 2015-03-19 2015-06-10 神华集团有限责任公司 基于NetWeaver的排程方法

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103853718B (zh) * 2012-11-28 2018-05-08 北京京东尚科信息技术有限公司 分片数据库访问方法及数据库系统
CN104133722B (zh) * 2014-07-02 2021-01-22 北京邮电大学 一种实现多业务流程协作的方法及系统
CN105550268B (zh) * 2015-12-10 2017-01-11 江苏曙光信息技术有限公司 大数据流程建模分析引擎
CN105787708A (zh) * 2016-03-16 2016-07-20 浪潮通用软件有限公司 一种基于任务池的单据处理方法
CN106920034B (zh) * 2017-02-07 2021-04-30 西南林业大学 一种挖掘bpmn编制流程并行性的方法及系统
CN109408134B (zh) * 2017-08-16 2022-04-08 阿里巴巴集团控股有限公司 模型文件的处理方法、装置、系统以及处理设备
CN109150957A (zh) * 2018-06-29 2019-01-04 中译语通科技股份有限公司 一种微服务并发控制系统
CN110377413B (zh) * 2019-07-24 2023-03-03 上海金融期货信息技术有限公司 基于bpmn标准的分布式任务异步调度和监控的系统
CN110704183B (zh) * 2019-09-18 2021-01-08 深圳前海大数金融服务有限公司 数据处理方法、系统及计算机可读存储介质
CN111585793B (zh) * 2020-04-20 2021-04-30 南京大学 一种网络服务优化组合方法
CN112527251A (zh) * 2020-11-30 2021-03-19 北京百度网讯科技有限公司 业务处理方法、装置、电子设备及存储介质
CN112540813B (zh) * 2021-01-09 2022-10-21 浙江工业大学 一种基于工作流引擎的应用生成方法
CN114915622B (zh) * 2022-05-31 2024-01-30 广东三维睿新科技有限公司 一种web端基于http的文件传输方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070260929A1 (en) * 2006-03-22 2007-11-08 Liu Peter T Pluggable debugging interface for graphical notation diagrams
CN101668047A (zh) * 2009-09-30 2010-03-10 北京航空航天大学 组合服务描述语言自动生成方法及装置
CN101695080A (zh) * 2009-09-30 2010-04-14 北京航空航天大学 Bpmn组合服务的执行方法和执行装置

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070260929A1 (en) * 2006-03-22 2007-11-08 Liu Peter T Pluggable debugging interface for graphical notation diagrams
CN101668047A (zh) * 2009-09-30 2010-03-10 北京航空航天大学 组合服务描述语言自动生成方法及装置
CN101695080A (zh) * 2009-09-30 2010-04-14 北京航空航天大学 Bpmn组合服务的执行方法和执行装置

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104699528A (zh) * 2015-03-19 2015-06-10 神华集团有限责任公司 基于NetWeaver的排程方法
CN104699528B (zh) * 2015-03-19 2018-05-04 神华集团有限责任公司 基于NetWeaver的排程方法

Also Published As

Publication number Publication date
CN102650953A (zh) 2012-08-29
CN102650953B (zh) 2014-05-07

Similar Documents

Publication Publication Date Title
WO2012116513A1 (zh) 并发优化的bpmn组合服务执行引擎及方法
CN108920259B (zh) 深度学习作业调度方法、系统和相关设备
WO2020108303A1 (zh) 一种基于异构计算的任务处理方法及软硬件框架系统
US8676848B2 (en) Configuring cloud resources
US8073671B2 (en) Dynamic software performance models
US9262210B2 (en) Light weight workload management server integration
Lipari et al. A framework for hierarchical scheduling on multiprocessors: from application requirements to run-time allocation
TWI280511B (en) Method for creating service instances in a computing grid, system thereof, and machine readable storage thereof
US20090113379A1 (en) Modeling and managing heterogeneous applications
US8682998B2 (en) Method and server cluster for map reducing flow services and large documents
CN110221901A (zh) 容器资源创建方法、装置、设备及计算机可读存储介质
CN109614227B (zh) 任务资源调配方法、装置、电子设备及计算机可读介质
CN104618433A (zh) 一种基于eca规则的服务组合方法
US10803413B1 (en) Workflow service with translator
Vukojevic-Haupt et al. On-demand provisioning of infrastructure, middleware and services for simulation workflows
Mathiya et al. Apache hadoop yarn parameter configuration challenges and optimization
CN108563509A (zh) 数据查询实现方法、装置、介质及电子设备
Al-Sinayyid et al. Job scheduler for streaming applications in heterogeneous distributed processing systems
Ghorbian et al. A survey on the scheduling mechanisms in serverless computing: a taxonomy, challenges, and trends
Hahn et al. TraDE-a transparent data exchange middleware for service choreographies
Rak et al. Chase: An autonomic service engine for cloud environments
Li et al. SoDa: A Serverless‐Oriented Deadline‐Aware Workflow Scheduling Engine for IoT Applications in Edge Clouds
US12086141B1 (en) Coordination of services using PartiQL queries
CN110968434A (zh) 一种管理机器的方法及相关装置
CN115242786B (zh) 一种基于容器集群的多模式大数据作业调度系统及方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 11859739

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 11859739

Country of ref document: EP

Kind code of ref document: A1