Running in production
To successfully run a distributed application in production, we need to consider a few more aspects beyond the best practices and patterns presented in the preceding sections. One specific area that comes to mind is introspection and monitoring. Let's go through the most important aspects in detail.
Logging
Once a distributed application is in production, it is not possible to debug it. But how can we then find out what exactly is the root cause of a malfunction of the application that has been reported by a user? The solution to this problem is to produce abundant and meaningful logging information. Developers need to instrument their application services in such a way that they output helpful information, such as when an error happens or a potentially unexpected or unwanted situation is encountered. Often, this information is output to STDOUT and STDERR, from where it is then collected by system daemons that write the information to local files or forward it to a central...
