A process manager is one of many failsafes used in case of a crash, and it also has many uses other than recovering from a crash. A robust production system sho

A process manager is one of many failsafes used in case of a crash, and it also has many uses other than recovering from a crash. A robust production system should have many layers of failsafes to prevent it from crashing, to recover quickly from crashes (increased load, outages, etc...), and to review crashes after-the-fact. In fact, most setup articles run every piece of software you mentioned in a process manager. They don't do this because they think the software will crash; they do it because when it does crash, they don't want to have to wait for a developer to manually reboot it while they diagnose the issue. Perhaps this is a good blog post topic.