SELECTIVELY RE-HOMING DATA PROCESSING TASKS BASED ON DATA COMPLEXITY AND CANDIDATE SYSTEM PERFORMANCE CHARACTERISTICS

摘要

Techniques are disclosed for selectively re-homing data processing tasks based on data complexity and candidate system performance characteristics. A server system obtains performance capabilities of a target system such that the server can decide whether to send program code and data to the target system for execution, based in part on the complexity of the data and code used in performing the request. In cases where a client system sending a request is capable of executing code and where other constraints are satisfied, e.g., bandwidth, throughput, and data volume, then the server can decide to re-home both the data and the program code to the client for execution.

主权项

1. A computer-implemented method for responding to a data processing request received by a first computing system, the method comprising:
obtaining a data complexity score corresponding to data used to perform the data processing request; obtaining a code complexity score corresponding to one or more code components executed to perform the data processing request; and determining, based on the data complexity score, code complexity score, and on performance metrics associated with a second computing system, whether to re-home the data and the one or more code components to the second computing system.