Abstract: An infrastructure system facilitates trace-based surveys to query users their experiences with pull request workflows. For example, the infrastructure system may receive from a client device associated with a user a pull request for making one or more changes to a programming code repository. The infrastructure system may generate a trace based on a pull request workflow including operations for making the change(s). The trace includes spans, each of which correspond to one or more operations in the pull request workflow. The infrastructure system generates log files, each log file includes data logged during the one or more operations of a span. The log files may be stored as a dataset of the trace. The infrastructure system may transmit a survey to the client device based on the dataset. The infrastructure system may also modify the pull request workflow based on the user's response to the survey.

Abstract: An infrastructure system facilitates trace-based surveys to query users their experiences with pull request workflows. For example, the infrastructure system may receive from a client device associated with a user a pull request for making one or more changes to a programming code repository. The infrastructure system may generate a trace based on a pull request workflow including operations for making the change(s). The trace includes spans, each of which correspond to one or more operations in the pull request workflow. The infrastructure system generates log files, each log file includes data logged during the one or more operations of a span. The log files may be stored as a dataset of the trace. The infrastructure system may transmit a survey to the client device based on the dataset. The infrastructure system may also modify the pull request workflow based on the user's response to the survey.

Abstract: Failures can occur when a software process is running, or when the software process is undergoing tests. A remote debugging session can be created to run and inspect the software process, so that a developer can fix the bug that caused a failure. Remote debugging sessions can consume valuable resources on cluster infrastructure. Debugging the failures can take up a significant amount of developer time. To alleviate some of these problems, a remote debugging session can be created on-demand at a later time when a developer is ready to start the remote debugging session. The deferred creation of the remote debugging session can be triggered by activating a uniform resource identifier that has information for creating the remote debugging session encoded therein. The resulting implementation makes it easy for a developer to create a remote debugging session, and resources can be allocated on an as-needed basis.