Now this is a root cause with transparency, and forward thinking resolution.
It seems that apache was having an unusual issue we've been seeing recently with DA and apache, where it fails to have a valid number of workers, and locks up and can't handle requests. I have added a reference configuration file with our general recommendations for MPM_Event, and then rebuilt the configurations in use to use these settings, and restarted apache in the process. This has fixed the problem in our experience, and should prevent it from happening again.