Processing element management in a streaming data system

• US 10,296,386 B2
• Filed: 11/20/2017
• Issued: 05/21/2019
• Est. Priority Date: 01/30/2012
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Branson, Michael J., Cradick, Ryan K., Santosuosso, John M., Schulz, Brandon W.
• Primary Examiner(s): Chea, Philip J
• Assistant Examiner(s): Ahmed, Mohammed
• Application Number:
  US15/817,426
• Publication Number:
  US 20180074862A1
• Time in Patent Office: 547 Days
• CPC Class Codes: G06F 2209/502 Proximity
• CPC Class Codes: G06F 9/44 Arrangements for executing specific programs
• CPC Class Codes: G06F 9/4881 Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
• CPC Class Codes: G06F 9/5011 the resources being hardware resources other than CPUs, Servers and Terminals
• CPC Class Codes: G06F 9/505 considering the load

Abstract

Stream applications may inefficiently use the hardware resources that execute the processing elements of the data stream. For example, a compute node may host four processing elements and execute each using a CPU. However, other CPUs on the compute node may sit idle. To take advantage of these available hardware resources, a stream programmer may identify one or more processing elements that may be cloned. The cloned processing elements may be used to generate a different execution path that is parallel to the execution path that includes the original processing elements. Because the cloned processing elements contain the same operators as the original processing elements, the data stream that was previously flowing through only the original processing element may be split and sent through both the original and cloned processing elements. In this manner, the parallel execution path may use underutilized hardware resources to increase the throughput of the data stream.

Claims

1. A method, comprising:

receiving streaming data to be processed by a plurality of processing elements comprising one or more of a plurality of operators;

establishing an operator graph defining at least one execution path in which a first operator of the plurality of operators is configured to receive data tuples from at least one upstream operator and transmit data tuples to at least one downstream operator;

processing the received streaming data using the first operator when the first operator is fused to a first processing element of the plurality of processing elements;

un-fusing the first operator from the first processing element;

before un-fusing the first operator, processing data within the first processing element using the first operator; and

transferring the first operator to a second processing element of the plurality of processing elements such that the first operator uses a first underutilized hardware resource to process the received streaming data.

2. The method of claim 1, further comprising, before transferring the first operator:

identifying, relative to predefined criteria, a second underutilized hardware resource in a computing system that executes the operators;

determining respective communication speeds between a hardware resource hosting the first operator and each of the first and second underutilized hardware resources; and

selecting, based on the respective communication speeds, one of the first and second underutilized hardware resources to host the first operator.

3. The method of claim 1, further comprising:

creating the second processing element for the first operator;

assigning the second processing element to the first underutilized hardware resource; and

determining whether the second processing element should be placed upstream or downstream from the first processing element in the operator graph to maintain the execution path.

4. The method of claim 1, wherein the first underutilized hardware resource is at least one of:

a CPU and a portion of memory.

5. The method of claim 4, further comprising:

identifying, relative to a predefined criteria, the first underutilized hardware resource in a computing system that executes the plurality of operations, wherein the first underutilized hardware resource is a CPU that is not executing any of the plurality of processing elements when the first underutilized hardware resource is identified.

6. The method of claim 1, further comprising:

identifying, relative to a predefined criteria, a second underutilized hardware resource in a computing system that executes the plurality of operations;

un-fusing a second operator from the first processing element of the plurality of processing elements;

transferring the second operator to a third processing element of the plurality of processing elements, wherein the third processing element is directly connected to the second processing element in the execution path; and

after transferring the second operator, processing at least a portion of the received streaming data using the second operator, wherein the second operator processes the portion of the received streaming data using the second underutilized hardware resource.

7. The method of claim 1, further comprising determining whether the first operator is un-fusable from the first processing element, wherein the first operator is un-fusable from the first processing element because the first processing element comprises more than one operator.

8. A computer program product comprising:

a computer-readable storage medium having computer-readable program code embodied therewith, the computer-readable program code comprising computer-readable program code configured to;

receive streaming data to be processed by a plurality of processing elements comprising one or more of a plurality of operators;

establish an operator graph defining at least one execution path in which a first operator of the plurality of operators is configured to receive data tuples from at least one upstream operator and transmit data tuples to at least one downstream operator;

process the received streaming data using the first operator when the first operator is fused to a first processing element of the plurality of processing elements;

un-fuse the first operator from the first processing element;

before un-fusing the first operator, process data within the first processing element using the first operator; and

transfer the first operator to a second processing element of the plurality of processing elements such that the first operator uses a first underutilized hardware resource to process the received streaming data.

9. The computer program product of claim 8, further comprising computer-readable program code configured to:

identify, relative to predefined criteria, a second underutilized hardware resource in a computing system that executes the operators;

determine respective communication speeds between a hardware resource hosting the first operator and each of the first and second underutilized hardware resources; and

select, based on the respective communication speeds, one of the first and second underutilized hardware resources to host the first operator.

10. The computer program product of claim 8, further comprising computer-readable program code configured to:

create the second processing element for the first operator;

assign the second processing element to the first underutilized hardware resource; and

determine whether the second processing element should be placed upstream or downstream from the first processing element in the operator graph to maintain the execution path.

11. The computer program product of claim 8, wherein the first underutilized hardware resource is at least one of:

a CPU and a portion of memory.

12. The computer program product of claim 11, wherein the first underutilized hardware resource is a CPU that is not executing any of the plurality of processing elements when the first underutilized hardware resource is identified.

13. The computer program product of claim 8, further comprising computer-readable program code configured to:

identify, relative to predefined criteria, a second underutilized hardware resource in the computing system;

un-fuse a second operator from the first processing element of the plurality of processing elements;

transfer the second operator into a third processing element of the plurality of processing elements, wherein the third processing element is directly connected to the second processing element in the execution path; and

after transferring the second operator, process at least a portion of the received streaming data using the second operator, wherein the second operator processes the portion of the received streaming data using the second underutilized hardware resource.

14. The computer program product of claim 8, further comprising computer-readable program code configured to determine whether the first operator is un-fusable from the first processing element, wherein the first operator is un-fusable because the first processing element comprises more than one operator.

15. A system, comprising:

a computer processor; and

a memory containing a program that, when executed on the computer processor, performs an operation for processing data, the operation comprising;

receiving streaming data to be processed by a plurality of processing elements comprising one or more of a plurality of operators;

establishing an operator graph defining at least one execution path in which a first operator of the plurality of operators is configured to receive data tuples from at least one upstream operator and transmit data tuples to at least one downstream operator;

processing the received streaming data using the first operator when the first operator is fused to a first processing element of the plurality of processing elements;

un-fusing the first operator from the first processing element;

before un-fusing the first operator, processing data within the first processing element using the first operator; and

transferring the first operator to a second processing element of the plurality of processing elements such that the first operator uses a first underutilized hardware resource to process the received streaming data.

16. The system of claim 15, wherein the operation further comprises:

identifying, relative to predefined criteria, a second underutilized hardware resource in a computing system that executes the operators;

determining respective communication speeds between a hardware resource hosting the first operator and each of the first and second underutilized hardware resources;

and selecting, based on the respective communication speeds, one of the first and second underutilized hardware resources to host the first operator.

17. The system of claim 15, wherein the operation further comprises:

creating the second processing element for the first operator;

assigning the second processing element to the first underutilized hardware resource; and

determining whether the second processing element should be placed upstream or downstream from the first processing element in the operator graph to maintain the execution path.

18. The system of claim 15, wherein the first underutilized hardware resource is at least one of:

a CPU and a portion of memory.

19. The system of claim 18, wherein the first underutilized hardware resource is a CPU that is not executing any of the plurality of processing elements when the first underutilized hardware resource is identified.

20. The system of claim 15, further comprising:

identifying, relative to predefined criteria, a second underutilized hardware resource in the computing system;

un-fusing a second operator from the first processing element of the plurality of processing elements;

transferring the second operator into a third processing element of the plurality of processing elements, wherein the third processing element is directly connected to the second processing element in the execution path; and

after transferring the second operator, processing at least a portion of the received streaming data using the second operator, wherein the second operator processes the portion of the received streaming data using the second underutilized hardware resource.