Capturing trace information using annotated trace output

摘要

A computer identifies a first trace output, generated during a first execution of a first program, that is annotated with an indication of a first pattern of logged events and one or more second programs that identify additional logged events. The computer identifies the first pattern of logged events in a second trace output, which is generated during a second execution of the first program. The computer executes the one or more second programs to gather one or more additional logged events that are discoverable during the second execution of the first program, wherein the one or more additional logged events are not included in the first trace output.

主权项

1. A method of capturing event information, the method comprising:
a computer accessing a first trace output, generated during a first execution of a first program; the computer annotating the first trace output with an indication of a first pattern of logged events that indicate a suspected error condition; the computer annotating the first trace output with and one or more second programs that identify additional logged events; the computer accessing a second trace output, generated during a second execution of the first program, the second trace output having a second pattern of logged events; the computer monitoring the second pattern of logged events for a match to the first pattern of logged events; and responsive to an identification of a match of the first pattern of logged events to the second pattern of logged events, the computer executing the one or more second programs to gather one or more additional logged events that are discoverable during the second execution of the first program, wherein the one or more additional logged events are not included in the first trace output.