Reconfigurable cloud computing
A method, system, and computer-readable storage medium for using a reconfigurable computing system are disclosed. For example, one method involves identifying a first application to be executed, configuring a computing node in a first configuration, and executing, according to a first workflow, at least a portion of a first application and at least another portion of the first application. The method also involves identifying a second application to be executed, configuring the computing node in a second configuration, and executing, according to a second workflow, at least a portion of a second application and at least another portion of the second application.
1. A method comprising:
identifying a first application to be executed;
configuring a computing node in a first configuration, wherein
the computing node comprises a plurality of computing resources,
the configuring the computing node in the first configuration comprises
configuring a first computing resource of the plurality of computing resources to execute at least a portion of the first application according to a first workflow, and
configuring a second computing resource of the plurality of computing resources to execute at least another portion of the first application according to the first workflow,
the first computing resource is of a first type of a plurality of types of computing resources, and
the second computing resource is of a second type of the plurality of types of computing resources;
executing, according to the first workflow, the at least the portion of the first application and the at least the another portion of the first application;
identifying a second application to be executed;
configuring the computing node in a second configuration, wherein
the configuring the computing node in the second configuration comprises
configuring the first computing resource to execute at least a portion of the second application according to a second workflow, and
configuring the second computing resource to execute at least another portion of the second application according to the second workflow; and
executing, according to the second workflow, the at least the portion of the second application and the at least the another portion of the second application.
2. The method of claim 1 , wherein
the configuring the computing node in the first configuration further comprises
selecting a first computing resource of the plurality of computing resources, and
selecting a second computing resource of the plurality of computing resources, and
the configuring the computing node in the second configuration further comprises
selecting the first computing resource, and
selecting the second computing resource.
3. The method of claim 2 , wherein
the first workflow and the second workflow are different portions of a single workflow.
4. The method of claim 2 , wherein
receiving the at least the portion of the first workflow at the computing node, wherein
the at least the portion of the first workflow comprises
a first plurality of scripts;
accessing a first configuration file, wherein
the first plurality of scripts references the first configuration file, and
the first computing resource is configured in the first configuration based on configuration information in the first configuration file;
receiving the at least the portion of the second workflow at the computing node, wherein
the at least the portion of the first workflow comprises
a second plurality of scripts; and
accessing a second configuration file, wherein
the second plurality of scripts references the second configuration file, and
the second computing resource is configured in the second configuration based on configuration information in the second configuration file.
5. The method of claim 4 , further comprising:
identifying a first job script, wherein
the first plurality of scripts comprises the first job script, and
the first job script identifies the at least the portion of the first application, the first computing resource, and a first configuration script;
identifying a second job script, wherein
the second plurality of scripts comprises the second job script, and
the second job script identifies the at least the portion of the second application, the first computing resource, and a second configuration script; and
accessing at least one control script, wherein
the at least one control script indicates that the first job script is to be executed before the second job script,
the at least one control script indicates that the first computing resource is to be configured in the first configuration prior to executing the at least the portion of the first application, and
the at least one control script indicates that the first computing resource is to be configured in the second configuration prior to executing the at least the portion of the second application.
6. The method of claim 1 , wherein
the first computing resource comprises a reconfigurable logic device, and
the first computing resource is selected to execute the at least the portion of the first application by
determining that one or more attributes of the reconfigurable logic device satisfy a resource requirement for the at least the portion of the first application, and
determining that the reconfigurable logic device can be configured to execute the at least the portion of the first application according to the first workflow, prior to executing the at least the portion of the first application.
7. The method of claim 6 , wherein
the reconfigurable logic device is a field-programmable gate array (FPGA), and
the configuring the first computing resource according to the first workflow comprises
changing a bit stream of the FPGA.
8. The method of claim 1 , wherein
the first computing resource comprises a graphics processing unit (GPU), and
the first computing resource is selected to execute the at least the portion of the first application by
determining that one or more attributes of the GPU satisfy a resource requirement for the at least the portion of the first application, and
determining that the GPU can be configured to execute the at least the portion of the first application according to the first workflow, prior to executing the at least the portion of the first application.
9. The method of claim 8 , wherein
the configuring the first computing resource according to the first workflow comprises
reconfiguring one or more registers of the GPU and/or changing an operation mode of the GPU.
10. The method of claim 1 , wherein
the plurality of types of computing resources comprises at least one of
a field-programmable gate array (FPGA),
a graphics processing unit (GPU),
a digital signal processor (DSP), or
a central processing unit (CPU).
11. The method of claim 1 , wherein
the at least the portion of the first application and the at least another portion of the first application are different portions of the first application.
12. A system comprising:
a computing node, wherein
the computing node comprises
a plurality of computing resources,
a computer-readable storage medium, coupled to the plurality of computing resources,
a configurator, stored in the computer-readable storage medium, and
a resource manager, stored in the computer-readable storage medium,
the configurator is configured to
identify a first application to be executed,
configure the computing node in a first configuration by virtue of being configured to
configure a first computing resource of the plurality of computing resources to execute at least a portion of the first application according to a first workflow, and
configure a second computing resource of the plurality of computing resources to execute at least another portion of the first application according to the first workflow,
the first computing resource is of a first type of a plurality of types of computing resources, and
the second computing resource is of a second type of the plurality of types of computing resources,
the resource manager is configured to
cause execution, according to the first workflow, of the at least the portion of the first application and the at least the another portion of the first application,
the configurator is further configured to
identify a second application to be executed, and
configure the computing node in a second configuration, wherein
the computing node is configured in the second configuration by virtue of being configured to
configure the first computing resource to execute at least a portion of the second application according to a second workflow,
configure the second computing resource to execute at least another portion of the second application according to the second workflow, and
the resource manager is further configured to
cause execution of the at least the portion of the second application and the at least the another portion of the second application.
13. The system of claim 12 , wherein
the configurator is further configured to
configure the computing node in the first configuration by virtue of being further configured to
select the first computing resource, and
select the second computing resource, and
configure the computing node in the second configuration by virtue of being further configured to
select the first computing resource, and
select the second computing resource.
14. The system of claim 13 , wherein the computing node is configured to:
receive the at least the portion of the first workflow at the computing node, wherein
the at least the portion of the first workflow comprises
a first plurality of scripts;
access a first configuration file, wherein
the first plurality of scripts references the first configuration file, and
the first computing resource is configured in the first configuration based on configuration information in the first configuration file;
receive the at least the portion of the second workflow at the computing node, wherein
the at least the portion of the second workflow comprises
a second plurality of scripts;
access a second configuration file, wherein
the second plurality of scripts references the second configuration file, and
the second computing resource is configured in the second configuration based on configuration information in the second configuration file;
identify a first job script, wherein
the first plurality of scripts comprises the first job script, and
the first job script identifies the at least the portion of the first application, the first computing resource, and a first configuration script;
identify a second job script, wherein
the second plurality of scripts comprises the second job script, and
the second job script identifies at least one portion of the second application, the first computing resource, and a second configuration script; and
access at least one control script, wherein
the at least one control script indicates that the first job script is to be executed before the second job script,
the at least one control script indicates that the first computing resource is to be configured in the first configuration prior to executing the at least one portion of the first application, and
the at least one control script indicates that the first computing resource is to be configured in the second configuration prior to executing the at least one portion of the second application.
15. The system of claim 12 , wherein
the first computing resource comprises a reconfigurable logic device, and
the first computing resource is selected to execute the at least the portion of the first application by
determining that one or more attributes of the reconfigurable logic device satisfy a resource requirement for the at least the portion of the first application, and
determining that the reconfigurable logic device can be configured to execute the at least the portion of the first application according to the first workflow, prior to executing the at least the portion of the first application.
16. The system of claim 15 , wherein
the reconfigurable logic device is a field-programmable gate array (FPGA), and
the configurator is configured to configure the first computing resource according to the first workflow by virtue of being configured to
change a bit stream of the FPGA.
17. The system of claim 12 , wherein
the first computing resource comprises a graphics processing units (GPU),
the GPU is selected to execute at least one portion of the first application by
determining that attributes of the GPU satisfy a resource requirement for the at least one portion of the first application, and
determining that the GPU can be configured in the first configuration prior to executing the at least one portion of the first application, and
the configurator is configured to configure the GPU in the second configuration by reconfiguring one or more registers of the GPU and/or changing an operation mode of the GPU.
18. A computer program product comprising:
a plurality of instructions, that, when executed by a hardware processor of a computer system, are configured to cause the hardware processor to identify a first application to be executed,
configure a computing node in a first configuration, wherein
the computing node comprises a plurality of computing resources,
the instructions configured to cause the hardware processor to configure the computing node in the first configuration comprise a plurality of instructions configured to
configure a first computing resource of the plurality of computing resources to execute at least a portion of the first application according to a first workflow, and
configure a second computing resource of the plurality of computing resources to execute at least another portion of the first application according to the first workflow,
the first computing resource is of a first type of a plurality of types of computing resources, and
the second computing resource is of a second type of the plurality of types of computing resources,
execute, according to the first workflow, the at least the portion of the first application and the at least the another portion of the first application,
identify a second application to be executed,
configure the computing node in a second configuration by virtue of being further configured to
configure the first computing resource to execute at least a portion of the second application according to a second workflow, and
configure the second computing resource to execute at least another portion of the second application according to the second workflow
execute, according to the second workflow, the at least the portion of the second application and the at least the another portion of the second application; and
a non-transitory computer-readable storage medium, wherein the instructions are encoded in the non-transitory computer-readable storage medium.
19. The computer program product of claim 18 , wherein
the plurality of instructions are further configured to configure the computing node in the first configuration are further configured to
select a first computing resource of the plurality of computing resources, and
select a second computing resource of the plurality of computing resources, and
the plurality of instructions are further configured to configure the computing node in the second configuration are further configured to
select the first computing resource, and
select the second computing resource.
20. The computer program product of claim 19 ,
wherein the plurality of instructions are further configured to:
receive the at least the portion of the first workflow at the computing node, wherein
the at least the portion of the first workflow comprises
a first plurality of scripts;
access a first configuration file, wherein
the first plurality of scripts references the first configuration file, and
the first computing resource is configured in the first configuration based on configuration information in the first configuration file;
receive the at least the portion of the second workflow at the computing node, wherein
the at least the portion of the first workflow comprises
a second plurality of scripts;
access a second configuration file, wherein
the second plurality of scripts references the second configuration file, and
the second computing resource is configured in the second configuration based on configuration information in the second configuration file;
identify a first job script, wherein
the first plurality of scripts comprises the first job script, and
the first job script identifies at least one portion of the first application, the first computing resource, and a first configuration script;
identify a second job script, wherein
the second plurality of scripts comprises the second job script, and
the second job script identifies the at least one portion of the second application, the first computing resource, and a second configuration script; and
access at least one control script, wherein
the at least one control script indicates that the first job script is to be executed before the second job script,
the at least one control script indicates that the first computing resource is to be configured in the first configuration prior to executing the at least one portion of the first application, and
the at least one control script indicates that the first computing resource is to be configured in the second configuration prior to executing the at least one portion of the second application.
21. The computer program product of claim 18 , wherein
the first computing resource comprises a reconfigurable logic device, and
the first computing resource is selected to execute the at least the portion of the first application by
determining that one or more attributes of the reconfigurable logic device satisfy a resource requirement for the at least the portion of the first application, and
determining that the reconfigurable logic device can be configured to execute the at least the portion of the first application according to the first workflow, prior to executing the at least the portion of the first application.
22. The computer program product of claim 18 , wherein
the first computing resource comprises a graphics processing unit (GPU), and
the first computing resource is selected to execute the at least the portion of the first application by
determining that one or more attributes of the GPU satisfy a resource requirement for the at least the portion of the first application, and
determining that the GPU can be configured to execute the at least the portion of the first application according to the first workflow, prior to executing the at least the portion of the first application.