No message when fsck is running at boot

Camaleón noelamac at gmail.com
Wed Aug 29 17:39:36 BST 2018


El 2018-08-15 a las 08:03 +0200, Camaleón escribió:

> El 2018-08-03 a las 17:24 +0200, Camaleón escribió:
> 
> > El 2018-08-03 a las 16:30 +0200, Andreas Henriksson escribió:
> > 
> > > On Sun, Jul 08, 2018 at 07:25:34PM +0200, Camaleón wrote:
> > > > Package: util-linux
> > > > Version: 2.29.2-1
> > > > 
> > > > On Debian Wheezy, when fsck was run at boot, it displayed a message and 
> > > > progress bar with percentage done.
> > > > 
> > > > Today, on Debian Strech, I just got an intermitent cursor and a black 
> > > > screen, no more no less, no other message was present, so I had to 
> > > > "reisub" to restart the system thinking something went wrong. This time 
> > > > I removed "quiet" from GRUB's menu and now fsck message and percentage 
> > > > was displayed as usual.
> > > > 
> > > > This is the first time this happens so it could have been a puntual 
> > > > glitch but I prefer to point this out and if needed, recap more 
> > > > information about it (not sure what logs to check).
> > > 
> > > This is not a problem with fsck itself, but whatever is launching it.
> > > In todays age the bootup phase is heavily parallelled. If you want
> > > to present anything meaningful to the user which isn't completely
> > > scrambled you need some kind of message aggregator. You can use
> > > plymouth for that as an example.
> > 
> > I will keep an eye over this. 
> > 
> > Should "fsck" (or any other program/task) is running in background and 
> > preventing/delaying system start up, a message should be displayed to 
> > tell the user, that's a must. No additional fancy boot splash should be 
> > needed for this at all (just think on non-GUI servers or non KMS enabled 
> > systems and so).
> 
> It just happened today. 
> 
> I'm attaching a snapshot for what user gets while fsck is on the go 
> (I mean... nothing).
> 
> Then, once the system boots, logs confirm the gap (10 minutes):
> 
> root at stt008:~# journalctl -u systemd-fsck*
> 
> -- Logs begin at Wed 2018-08-15 07:35:03 CEST, end at Wed 2018-08-15 07:50:35 CEST. --
> ago 15 07:35:06 stt008 systemd[1]: Starting File System Check on /dev/disk/by-uuid/8ab7cc61-8d69-42d6-bfb8-112f0e0f8a40...
> ago 15 07:35:06 stt008 systemd[1]: Started File System Check Daemon to report status.
> ago 15 07:35:07 stt008 systemd-fsck[305]: DATA ha sido montado 38 veces sin ser revisado, se fuerza la revisión.
> ago 15 07:45:43 stt008 systemd-fsck[305]: DATA: 222/30531584 ficheros (19.4% no contiguos), 109375618/122096000 bloques
> ago 15 07:45:44 stt008 systemd[1]: Started File System Check on /dev/disk/by-uuid/8ab7cc61-8d69-42d6-bfb8-112f0e0f8a40.
> 
> Again, a message (or whatever) displaying fsck is running and showing the 
> current status/percentage is desiderable.

Adding more information. 

Maybe related to bug #775093 so adding systemd package maintainers to 
shed some light over this (if possible).

Also, the behaviour seems to happened on Jessie:

jessie -- systemd fsck looks like computer is hanging
https://www.linuxquestions.org/questions/debian-26/jessie-systemd-fsck-looks-like-computer-is-hanging-4175513749/

Greetings,

-- 
Camaleón 



More information about the Pkg-systemd-maintainers mailing list