Open jacdavi opened 1 year ago
Some additional context. When running concurrently against multiple tenants our Write-Warning/Write-Error log messages become intermixed with other event messages. It's also hard to distinguish which tenant ScubaGear is running against that is causing the error messages to pop up. The above enhancement would rectify this.
Update. Throwing a non-zero exit code on all scenarios without stubbing ScubaGear's error handling toes is quite difficult. Will start looking into adding error handling flags within the JSON itself for a programmatic CRC like check.
💡 Summary
Currently it seems ScubaGear exits with a 0 code regardless of any errors that occur during runtime.
Motivation and context
This would help with identifying issues during automated runs. e.g., a non-zero exit code triggers some type of notification for a user to check the report (which I believe will contain errors after #65 ). Otherwise, our only way to detect errors automatically would be to parse the stderr output or potentially parse the output files.
Acceptance criteria
How do we know when this work is done?