Abstract: Examples relate to load testing. The examples disclosed herein enable obtaining lines of code that are recorded as an application is executed in a client computing device, the lines of code being recorded in chronological order of the execution; determining whether a dependency on at least one variable exists in individual lines of the lines of code; in response to determining that the dependency exists, storing the dependency in a data storage; identifying, from the lines of code, a line of code including a network call statement that calls a called variable; and eliminating a first subset of the lines of code based on the called variable and dependencies stored in the data storage, wherein a second subset of the lines of code that remain after the elimination comprises user-entered parameter data.

Abstract: Example embodiments relate to generating sets of recommended inputs for changing predicted results of a predictive model. The examples disclosed herein access, from a database, a historical set of inputs and results of a predictive model. A function is approximated based on the historical set of inputs and results, and a gradient of the function is computed using a result of the function with respect to a local maximum value of the function. A set of recommended inputs is generated based on the gradient of the function, where a recommended input produces a positive result of the function.

Abstract: Techniques to create and use cluster models to predict build failures are provided. In one aspect, clusters in a set of builds may be identified. The identified clusters may be used to create a model. The model may be used to predict causes of build failures. In another aspect, a failed build may be identified. A clustering model may be retrieved. A cause of problems with the failed build may be predicted using the clustering model.

Abstract: Methods performed by a physical computing system include automatically identifying, using at least one trained classifier, an action for responding to an anomaly in the execution of the application in a cloud computing system. The at least one trained classifier relates a metrics set to a result of performing an action for addressing an anomaly. Systems and computer readable media are also described herein.

Abstract: Application management based on data correlations is disclosed. One example is a system including a data processor, a data element generator, a matrix generator, a data analysis module, a performance module, and a load test manager. The data processor accesses test data based on an application under load testing. The data element generator generates a plurality of transactional data elements based on the test data, each data element comprising at least three data components. The matrix generator generates a covariance matrix based on the data components. The data analysis module determines an eigenvector associated with the covariance matrix, and identifies a correlation between a sub-plurality of the at least three data components based on coefficients of the eigenvector. The performance module determines, based on the correlation, performance metrics for the application under load testing. The load test manager manages, based on the performance metrics, the application under load testing.

Abstract: Example embodiments relate to determining code coverage based on production sampling. In example embodiments, a production execution data set that includes metrics for code units of a software application is obtained, where the metrics include input and output values for each of the code units and an average execution count for each of the code units. Further, application code execution is tracked during a testing procedure of the software application to determine executed lines of code. At this stage, production code coverage of the software application is determined based on the production execution data set and the executed lines of code.

Abstract: Example embodiments relate to generating sets of recommended inputs for changing predicted results of a predictive model. The examples disclosed herein access, from a database, a historical set of inputs and results of a predictive model. A function is approximated based on the historical set of inputs and results, and a gradient of the function is computed using a result of the function with respect to a local maximum value of the function. A set of recommended inputs is generated based on the gradient of the function, where a recommended input produces a positive result of the function.

Abstract: Techniques for identifying a build commit that caused a test failure are provided. A build which includes a failed test may be identified. For each commit in the build a weighting factor may be calculated for files that have been previously associated with the failed test. The weighting factor may be based on the number of times the file has been associated with the failed test and the total number of tests. A weighting factor may also be calculated for files that have not been previously associated with the failed test based on the number of times the file appears with other files that are associated with the failed test. The weighting factors may be added to create a score for the commit. The scores for the commits in the build may be ordered. The higher the score, the more likely the commit was the cause of the failed test.

Abstract: Techniques to create and use cluster models to predict build failures are provided. In one aspect, clusters in a set of builds may be identified. The identified clusters may be used to create a model. The model may be used to predict causes of build failures. In another aspect, a failed build may be identified. A clustering model may be retrieved. A cause of problems with the failed build may be predicted using the clustering model.

Abstract: Techniques for identifying a build commit that caused a test failure are provided. A build which includes a failed test may be identified. For each commit in the build a weighting factor may be calculated for files that have been previously associated with the failed test. The weighting factor may be based on the number of times the file has been associated with the failed test and the total number of tests. A weighting factor may also be calculated for files that have not been previously associated with the failed test based on the number of times the file appears with other files that are associated with the failed test. The weighting factors may be added to create a score for the commit. The scores for the commits in the build may be ordered. The higher the score, the more likely the commit was the cause of the failed test.

Abstract: In one example of the disclosure, code lines for a software program are received, the code lines including a unit of code lines. Code entities within the unit are identified. Each code entity includes a line or consecutive lines of code implementing a distinct program requirement or defect fix for the program. Context changes are identified within the unit, each context change including an occurrence of a first code line set implementing an entity, adjacent to a second code line set implementing another entity, within a same code scope. A code complexity score is determined based upon counts of entities identified and context changes identified within the unit, and upon counts of code lines and entities within the program.

Abstract: Systems, methods, and machine-readable and executable instructions are provided for error developer association. Error developer association can include identifying a number of portions of the source code associated with a message, wherein the message is associated with an error. Error developer association can also include associating a developer with a portion of the source code of the number of portions of the source code. Error developer association can also include identifying a developer of the number of developers to resolve the error.

Abstract: Examples relate to providing distributed compilation of statically typed languages. In some examples, first order dependencies of a target module are identified, where each of the first order dependencies is associated with one of a number of dependent modules. Next, each first order dependency of is traversed to remove code references from source code of a corresponding module of the plurality of dependent modules, where each of the code references refers to a type defined in an indirect dependency of the target module, and compile the source code of the corresponding module to generate a module stub of a number of module stubs. At this stage, source code of the target module is compiled using the module stubs to generate a target program.

Abstract: In one example of the disclosure, event notices are received, with each notice indicative of degradation of a configuration item. Configuration item past-preference pairings are accessed. Each pairing includes a count of operator-exhibited preferences for event notices associated with a first configuration item relative to event notices associated with a second configuration item. A prioritized ordering of the received event notices is created utilizing the past-preference pairings.

Abstract: A risk level of a software commit is assessed through the use of a classifier. The classifier may be generated based on attributes pertaining to previous commits and used to determine a risk level for deployment of a software commit into a production environment based on attributes extracted from the software commit.

Abstract: Application management based on data correlations is disclosed. One example is a system including a data processor, a data element generator, a matrix generator, a data analysis module, a performance module, and a load test manager. The data processor accesses test data based on an application under load testing. The data element generator generates a plurality of transactional data elements based on the test data, each data element comprising at least three data components. The matrix generator generates a covariance matrix based on the data components. The data analysis module determines an eigenvector associated with the covariance matrix, and identifies a correlation between a sub-plurality of the at least three data components based on coefficients of the eigenvector. The performance module determines, based on the correlation, performance metrics for the application under load testing. The load test manager manages, based on the performance metrics, the application under load testing.

Abstract: Examples relate to load testing. The examples disclosed herein enable obtaining lines of code that are recorded as an application is executed in a client computing device, the lines of code being recorded in chronological order of the execution; determining whether a dependency on at least one variable exists in individual lines of the lines of code; in response to determining that the dependency exists, storing the dependency in a data storage; identifying, from the lines of code, a line of code including a network call statement that calls a called variable; and eliminating a first subset of the lines of code based on the called variable and dependencies stored in the data storage, wherein a second subset of the lines of code that remain after the elimination comprises user-entered parameter data.

Abstract: In one example in accordance with the present disclosure, a method for test session similarity determination includes capturing a sequence of events from a user session of an application and converting the captured sequence into a data format used for a test sequence. The method also includes converting each event in the test sequence that is not in the captured sequence into a disparate event and creating a unique set including each unique event in the captured sequence and the disparate event. The method also includes determining a first average relative location of the event in the captured sequence and a second average relative location of each event in the rest sequence. The method also includes determining a degree of similarity between the captured sequence and the test sequence based on a comparison of the first and second average relative location and automatically generating a visualization highlighting the degree of similarity.

Abstract: Examples relate to providing distributed compilation of statically typed languages. In some examples, first order dependencies of a target module are identified, where each of the first order dependencies is associated with one of a number of dependent modules. Next, each first order dependency of is traversed to remove code references from source code of a corresponding module of the plurality of dependent modules, where each of the code references refers to a type defined in an indirect dependency of the target module, and compile the source code of the corresponding module to generate a module stub of a number of module stubs. At this stage, source code of the target module is compiled using the module stubs to generate a target program.

Abstract: In one example of the disclosure, code lines for a software program are received, the code lines including a unit of code lines. Code entities within the unit are identified. Each code entity includes a line or consecutive lines of code implementing a distinct program requirement or defect fix for the program. Context changes are identified within the unit, each context change including an occurrence of a first code line set implementing an entity, adjacent to a second code line set implementing another entity, within a same code scope. A code complexity score is determined based upon counts of entities identified and context changes identified within the unit, and upon counts of code lines and entities within the program.