On Sat, Feb 11 2017, Michael Biebl wrote: >> reached? If I cannot get a clean journal out, there's nothing much I can >> debug. > > You could try to attach a serial device to your VM and log the output to > that device [1]. This will get us a more complete log. > Attach the complete file.
But for reference, how would I proceed on real hardware instead? I had multiple occasions in the past with systemd+cryptsetup with where the root fs was left unsynced and the journal becomes corrupted. I couldn't find a way to reliable force an unit to be sequenced just prior to shutdown.target.
