Comment # 13 on bug 943766 from Cristian Rodr�guez

(In reply to Marvin FourtyTwo from comment #12)
> It means that after the first timeout (90 secs) it is enlarged to 180 secs
> and after that service is declared failed. Then systemd tries to start it
> again and timeout begins again for zero. I had a problem with my journal
> logs which were never written to disk. I'll try to get some more debug info
> soon

systemctl enable debug-shell.service ..then boot with "debug" and without
splash=silent quiet.. a root shell will start very early which you can use
pressing ctrl+alt f9.. then dmesg > debug.log