Evaluating and remediating source code variability

A computing system having a memory having computer readable instructions and one or more processors for executing the computer readable instructions, the computer readable instructions controlling the one or more processors to perform operations comprising:

• performing a static analysis on a source code to identify one or more functions of the source code and a type associated with each of the one or more functions;
• defining a threshold minimum number of times for execution of the one or more functions of the source code to collect an adequate data set of telemetry data for the one or more functions;
• performing a dynamic analysis on the source code by executing the source code and capturing telemetry data for the execution of each of the one or more functions, a state of the one or more processors before and after the execution of each of the one or more functions, and a state of the memory before and after the execution of each of the one or more functions and wherein the captured telemetry data includes an execution time associated with each of the one or more functions and wherein the dynamic analysis includes execution of the one or more functions of the source code the threshold minimum number of times;
• calculating a source code execution variability score for each of the one or more functions based at least in part on the execution time, the state of the one or more processors before and after execution, and the state of the memory before and after execution, and the type associated with each of the one or more functions;
• based on a determination that the source code execution variability score associated with one of the one or more functions is above a threshold value, wherein the threshold value is based at least in part on a network traffic value in a cloud computing environment, remediating the one of the one or more functions,
• wherein the remediation includes prompting a large language model to create an alternative version of the one or more functions, wherein the alternative version is generated to reduce the source code execution variability score, wherein the alternative version of the one or more functions generated by the large language model is validated by executing the alternative version and comparing its execution variability score to the threshold value;
• and
• updating the source code by replacing the one or more functions with the alternative version,
• wherein executing the source code includes executing the source code in the cloud computing environment and receiving application program interface (API) calls to perform the one or more functions.

A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to perform operations comprising:

• performing a static analysis on a source code to identify one or more functions of the source code and a type associated with each of the one or more functions;
• defining a threshold minimum number of times for execution of the one or more functions of the source code to collect an adequate data set of telemetry data for the one or more functions;
• performing a dynamic analysis on the source code by executing the source code and capturing telemetry data for the execution of each of the one or more functions, a state of the processor before and after the execution of each of the one or more functions, and a state of a memory before and after the execution of each of the one or more functions and wherein the captured telemetry data includes an execution time associated with each of the one or more functions and wherein the dynamic analysis includes execution of the one or more functions of the source code the threshold minimum number of times;
• calculating a source code execution variability score for each of the one or more functions based at least in part on the execution time, the state of the processor before and after execution, and the state of the memory before and after execution, and the type associated with each of the one or more functions;
• and
• based on a determination that the source code execution variability score associated with one of the one or more functions is above a threshold value, wherein the threshold value is based at least in part on a network traffic value in a cloud computing environment, remediating the one of the one or more functions,
• wherein the remediation includes prompting a large language model to create an alternative version of the one or more functions, wherein the alternative version is generated to reduce the source code execution variability score, wherein the alternative version of the one or more functions generated by the large language model is validated by executing the alternative version and comparing its execution variability score to the threshold value;
• and
• updating the source code by replacing the one or more functions with the alternative version,
• wherein executing the source code includes executing the source code in the cloud computing environment and receiving application program interface (API) calls to perform the one or more functions.