Improve logging - Make pipeline errors obvious to end user for subprocess exceptions
Coming out of a discussion with CHT, we want to ensure that the error messages are tailored towards the end-user rather than the pipeline engineer. Therefore, we want to improve these logs so that a container contributor finds it more clear how to resolve the error.
For example, in build stage when they can't access the internet, we have noticed that it is not clear to the user why the pipeline is failing and what changes they have to make.
The main work of this ticket is to catch the GenericSubprocess errors we're throwing, and update the logging for those to:
- No longer print the stack trace We should also update any other logging we deem confusing, or create another ticket to track other logging we need to improve.
- Provide helpful logging for different contextual situations
- Examples
- Unable to reach resource during build (this could be due to accessing a resource we're not supporting in nexus, or unable to get resource from nexus)
- Cosign verify failed
- Skopeo copy failed
- any others we can think of
- Examples
Edited by Harmanjit Bhullar