Designing for Debuggability

1. The Question You Cannot Answer Later

Debuggability is the property that a system can answer questions nobody wrote down in advance.

The test is specific. During an incident somebody asks: which of these requests were slow, and what did they have in common? A system designed for debuggability answers that in minutes. A system that was not answers it never, because the data needed to answer it was not recorded while the requests were happening.

That asymmetry is why this belongs in the design rather than in a later ticket.

• You can add a dashboard after an incident