Investigation in production
Once we are aware that we have a problem in production, we need to understand what is happening and what the key elements that produce it are.
It's very important to remark on the importance of being able to replicate a problem. If that's the case, tests can be done to produce the error and follow the consequences.
The most important tools when analyzing why a particular problem is produced are the observability tools. That's why it is important to do preparation work in advance to be sure to be able to find problems when required.
We talked in previous chapters about logs and metrics. When debugging, metrics are normally not relevant, except to show the relative importance of a bug. Checking an increase in returned errors can be important to detect that there's an error, but detecting what error will require more precise information.
Do not underestimate metrics, though. They can help quickly determine what...
