SELinux is disabled. [root@krom jeff]# service cherokee enable Redirecting to /bin/systemctl enable cherokee.service [root@krom jeff]# service cherokee start Redirecting to /bin/systemctl start cherokee.service (wait a minute or so) Job failed. See system journal and 'systemctl status' for details. [root@krom jeff]# service cherokee status Redirecting to /bin/systemctl status cherokee.service cherokee.service - Cherokee Lightweight Web Server Loaded: loaded (/usr/lib/systemd/system/cherokee.service; enabled) Active: failed (Result: timeout) since Fri, 10 Aug 2012 16:12:10 -0400 CGroup: name=systemd:/system/cherokee.service And yet, *during the minute where service cherokee start is running* (ie: before it supposedly spits a timeout error), cherokee is actually running fine (you can access the web server's pages and all that). It is only when systemd "gives up" that it kills the cherokee server and everything becomes inaccessible.
Solution https://bugzilla.redhat.com/show_bug.cgi?id=786748#c3
Works for me now in Fedora 18.