mpm_worker was notoriously bad back then and would frequently deadlock the entire Apache process. So the default was to either use prefork or some sort of monit

mpm_worker was notoriously bad back then and would frequently deadlock the entire Apache process. So the default was to either use prefork or some sort of monitoring to restart stuck processes