# US 10,769,180 B2 · Efficient dataflow processing for objects > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,769,180 B2 - Document ID: US10769180B2 - Status: Granted - Kind: B2 - Application number: 15/423,523 - Filed: 2017-02-02 - Published: 2018-08-02 - Granted: 2020-09-08 - Earliest claimed: 2017-02-02 - Assignees: International Business Machines Corporation - Applicants: International Business Machines Corporation - Inventors: Ronald J. Barber; Rene Mueller; Vijayshankar Raman; Richard S. Sidle; Pinar Tozun - Web page: https://patentgrep.com/patent/US10769180B2 ## CPC - G06F 16/289: {Object oriented databases} - G06F 9/4488: {Object-oriented} - G06F 9/451: Execution arrangements for user interfaces - G06F 9/45508: {Runtime interpretation or emulation, e g. emulator loops, bytecode interpretation} - G06F 9/46: Multiprogramming arrangements - G06F 11/14: Error detection or correction of the data by redundancy in operations (error detection or correction of the data by redundancy in hardware G06F11/16) - G06F 16/235: {Update request formulation} - G06F 16/288: {Entity relationship models} - G06F 8/41: Compilation - G06F 2201/80: Database-specific techniques - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract One embodiment provides a method comprising adjusting a runtime of a dataflow processing environment to operate on multiple batches of objects. The method further comprises pre-allocating one or more vectors of objects, and processing the multiple batches one at a time. The one or more vectors of objects are re-used during processing of each batch. ## Claims ### Claim 1 A method comprising: changing a runtime of a distributed parallel batch-based dataflow processing environment to operate on multiple batches of object-oriented application objects, wherein the changed runtime provides that each processing thread uses its own vector of object-oriented application objects to process its own batches of object-oriented application objects, and each object is an instance of a class that is nested containing other classes; pre-allocating one or more vectors of object-oriented application objects for nested classes a single time, wherein the one or more vectors are extendible; assigning the one or more vectors of object-oriented application objects to a processing thread that processes an object-oriented application; and processing the multiple batches one at a time by the processing thread, wherein the one or more vectors of object-oriented application objects are re-used multiple times during processing of each new batch of object-oriented application objects by the processing thread; wherein a string is implemented as a complex object, the string is nested in an object using a serialized format in which all strings of a batch are concatenated together into a growable array, and the processing thread completes processing each batch before a next batch enters a pipeline of the processing thread. ### Claim 2 (depends on 1) The method of claim 1, wherein each object-oriented application object of each batch is a complex object with one or more levels of nesting and re-use of the one or more vectors of object-oriented application objects eliminates memory allocation and garbage collection in the distributed parallel dataflow batch-based processing environment. ### Claim 3 (depends on 2) The method of claim 2, further comprising: scanning a class hierarchy of the complex object to determine a set of vectors to allocate with one vector per class. ### Claim 4 (depends on 3) The method of claim 3, wherein the allocated set of vectors is used as a common location to place object-oriented application objects during processing of each of the multiple batches, and content at each level of nesting of the complex object is placed in a vector allocated for a corresponding class in the class hierarchy. ### Claim 5 (depends on 2) The method of claim 2, wherein the complex object further comprises at least one of a commonly used object-oriented application object type, or an array, and data structures are created for the complex object such that object-oriented application object reuse is enabled in the distributed parallel batch-based dataflow processing environment with the changed runtime. ### Claim 6 (depends on 1) The method of claim 1, wherein the one or more vectors of object-oriented application objects are extendible, and if a subsequent batch of object-oriented application objects has more content than will fit in a currently allocated size of a vector of object-oriented application objects, the size of the vector of object-oriented application objects is extended such that the vector of object-oriented application objects can hold content of the subsequent batch of object-oriented application objects. ### Claim 7 (depends on 2) The method of claim 2, wherein: the complex object is an instance of a polymorphic class having different distinct sub-types; and the method further comprises: for each distinct sub-type: pre-allocating a corresponding separate vector of object-oriented application objects; and processing, per sub-type, a batch of object-oriented application objects having the distinct sub-type independently using the corresponding vector of object-oriented application objects; combining processing results for each vector of object-oriented application objects pre-allocated for each distinct sub-type; and virtual function calls are reduced in presence of polymorphism. ### Claim 8 A system comprising: at least one processor; and a non-transitory processor-readable memory device storing instructions that when executed by the at least one processor causes the at least one processor to perform operations including: changing a runtime of a distributed parallel batch-based dataflow processing environment to operate on multiple batches of object-oriented application objects, wherein the changed runtime provides that each processing thread uses its own vector of object-oriented application objects to process its own batches of object-oriented application objects, and each object is an instance of a class that is nested containing other classes; pre-allocating one or more vectors of object-oriented application objects for nested classes a single time; assigning the one or more vectors of object-oriented application objects to a processing thread that processes an object-oriented application; and processing the multiple batches one at a time by the processing thread, wherein the one or more vectors of object-oriented application objects are re-used multiple times during processing of each new batch of object-oriented application objects by the processing thread; wherein a string is implemented as a complex object, the string is nested in an object using a serialized format in which all strings of a batch are concatenated together into a growable array, and the processing thread completes processing each batch before a next batch enters a pipeline of the processing thread. ### Claim 9 (depends on 8) The system of claim 8, wherein each object-oriented application object of each batch is a complex object with one or more levels of nesting and re-use of the one or more vectors of object-oriented application objects eliminates memory allocation and garbage collection in the distributed parallel dataflow batch-based processing environment. ### Claim 10 (depends on 9) The system of claim 9, the operations further comprising: scanning a class hierarchy of the complex object to determine a set of vectors to allocate with one vector per class. ### Claim 11 (depends on 10) The system of claim 10, wherein the allocated set of vectors is used as a common location to place object-oriented application objects during processing of each of the multiple batches, and content at each level of nesting of the complex object is placed in a vector allocated for a corresponding class in the class hierarchy. ### Claim 12 (depends on 9) The system of claim 9, wherein the complex object further comprises at least one of a commonly used object-oriented application object type, or an array, and data structures are created for the complex object such that object-oriented application object reuse is enabled in the distributed parallel batch-based dataflow processing environment with the changed runtime. ### Claim 13 (depends on 8) The system of claim 8, wherein the one or more vectors of object-oriented application objects are extendible, and if a subsequent batch of object-oriented application objects has more content than will fit in a currently allocated size of a vector of object-oriented application objects, a size of the vector of object-oriented application objects is extended such that the vector of object-oriented application objects can hold content of the subsequent batch of object-oriented application objects. ### Claim 14 (depends on 9) The system of claim 9, wherein: the complex object is an instance of a polymorphic class having different distinct sub-types; and the operations further comprise: for each distinct sub-type: pre-allocating a corresponding separate vector of object-oriented application objects; and processing, per sub-type, a batch of object-oriented application objects having the distinct sub-type independently using the corresponding vector of object-oriented application objects; combining processing results for each vector of object-oriented application objects pre-allocated for each distinct sub-type; and virtual function calls are reduced in presence of polymorphism. ### Claim 15 A computer program product comprising a computer-readable hardware storage medium having program code embodied therewith, the program code being executable by a computer to implement a method comprising: changing a runtime of a distributed parallel batch-based dataflow processing environment to operate on multiple batches of object-oriented application objects, wherein the changed runtime provides that each processing thread uses its own vector of object-oriented application objects to process its own batches of object-oriented application objects, and each object is an instance of a class that is nested containing other classes; pre-allocating one or more vectors of object-oriented application objects for nested classes a single time; assigning the one or more vectors of object-oriented application objects to a processing thread that processes an object-oriented application; and processing the multiple batches one at a time by the processing thread, wherein the one or more vectors of object-oriented application objects are re-used multiple times during processing of each new batch of object-oriented application objects by the processing thread; wherein a string is implemented as a complex object, the string is nested in an object using a serialized format in which all strings of a batch are concatenated together into a growable array, and the processing thread completes processing each batch before a next batch enters a pipeline of the processing thread. ### Claim 16 (depends on 15) The computer program product of claim 15, wherein: each object-oriented application object of each batch is a complex object with one or more levels of nesting and re-use of the one or more vectors of object-oriented application objects eliminates memory allocation and garbage collection in the distributed parallel dataflow batch-based processing environment; the complex object further comprises at least one of a commonly used object-oriented application object type, or an array; and data structures are created for the complex object such that object-oriented application object reuse is enabled in the distributed parallel batch-based dataflow processing environment with the changed runtime. ### Claim 17 (depends on 16) The computer program product of claim 16, the method further comprising: scanning a class hierarchy of the complex object to determine a set of vectors to allocate with one vector per class. ### Claim 18 (depends on 17) The computer program product of claim 17, wherein the allocated set of vectors is used as a common location to place object-oriented application objects during processing of each of the multiple batches, and content at each level of nesting of the complex object is placed in a vector allocated for a corresponding class in the class hierarchy. ### Claim 19 (depends on 15) The computer program product of claim 15, wherein the one or more vectors of object-oriented application objects are extendible, and if a subsequent batch of object-oriented application objects has more content than will fit in a currently allocated size of a vector of object-oriented application objects, a size of the vector of object-oriented application objects is extended such that the vector of object-oriented application objects can hold content of the subsequent batch of object-oriented application objects. ### Claim 20 (depends on 16) The computer program product of claim 16, wherein: the complex object is an instance of a polymorphic class having different distinct sub-types; and the method further comprises: for each distinct sub-type: pre-allocating a corresponding separate vector of object-oriented application objects; and processing, per sub-type, a batch of object-oriented application objects having the distinct sub-type independently using the corresponding vector of object-oriented application objects; combining processing results for each vector of object-oriented application objects pre-allocated for each distinct sub-type; and virtual function calls are reduced in presence of polymorphism.