While it's no one's favorite subject (outside some very strange individuals I know, one of whom is a technical editor for this book), logging and monitoring is a crucial part of any installation, big or small.
You want to know when your boxes have died or, better yet, when they're about to die, and you also want to be able to retroactively work out why they conked out in the first place.
Monitoring and logging can be as complex or simple as you want it to be. Some companies hire specific individuals to deal with these components in a silo, but, in smaller organizations, it could very well be you who ends up managing and configuring everything. If this is the case, then I currently recommend setting up Icinga2 and some sort of Elastic Stack implementation, but your needs and budget may vary.
One elephant in the room that we need to...
