Is system 'logging' captured in user stories or are they part of implementation detail?

Question

I understand that anyone who supports and maintains system can be seen (or should be seen?) as a stakeholder/user.

For me, so far 'logging' has been like:

"I guess, it wouldn't hurt to log this activity here"

or

"This error definitely needs to be logged.".

I'm leaning towards that logging is a subtle part of implementation detail.

However. I can construct some stories.

As a maintainer,
I would like to read systems activities of problematic (payment)transaction,
so that I can find and fix the problem.

As a maintainer,
In case of unhandled error I would like to read potential systems activities that lead to that situation,
so that I can find and fix that problem.

Additionally, I have been in the situation where client encountered an error in production and told us to read the logs for details and I was like 'what logs?' - he took 'logging' for granted.
There were logs, but nevertheless, I was a bit surprised.

Is system 'logging' captured in user stories (treated as any other feature) or are they part of implementation detail?

Answer 1

I would not expect it to be captured in the story. I would expect logging standards to be required by the definition of done or, if there was some one-off logging need, I'd expect it to come up in the team's conversation during planning or refinement and maybe have a task under that story.

Answer 2

I would recommend that you create nonfunctional requirements in the form of user stories to provide the foundation or infrastructure for logging.

This will allow all future features to leverage this and allow logging to be part and parcel of every story and therefore part of the definition of done.