Cyril Brulebois <kibi@debian.org> (2015-02-20):
> Control: severity -1 important
>
> Axel Beckert <abe@debian.org> (2015-02-04):
> > Control: found -1 20150107
> > Control: severity -1 serious
> >
> > Hi,
> >
> > MARTON Jozsef wrote:
> > > The formatting dialog hanged at 33%.
> > >
> > > * What outcome did you expect instead?
> > >
> > > To re-format /dev/sda3 creating a fresh ext4 filesystem.
> > >
> > >
> > > After examining the situation and killing the mkfs.ext4 process on a
> > > second virtual terminal, the installer gave a red screen error and
> > > returned to the partitioning menu.
> > >
> > > Invoking mkfs.ext4 manually on the sencond VT, it gave the yes-no
> > > question that follows (captured under the running Jessie install, so
> > > mke2fs version might be different). This might be the point where
> > > the installer hanged at.
> > >
> > > # mkfs.ext4 /dev/sda3
> > > mke2fs 1.42.12 (29-Aug-2014)
> > > /dev/sda3 contains a ext4 file system
> > > last mounted on / on Sat Nov 1 20:18:44 2014
> > > Proceed anyway? (y,n)
> >
> > We ran into this today with the current Jessie installer (labeled RC1
> > IIRC). Context was the reinstallation of a computer with the same
> > partition layout.
> >
> > The question vanishes if a -F is passed to the mkfs.ext* call, so this
> > should be an easy fix:
> >
> > -F Force mke2fs to create a filesystem, even if the
> > specified device is not a partition on a block special
> > device, or if other parameters do not make sense. In
> > order to force mke2fs to create a filesystem even if the
> > filesystem appears to be in use or is mounted (a truly
> > dangerous thing to do), this option must be specified twice.
> >
> > > After formatting from the second VT, installer was able to continue
> > > without re-formatting the partition. An other solution was to zero
> > > out the forst few megabytes of sda3 using dd and /dev/zero to allow
> > > formatting by the installer.
> >
> > I consider both variants unsuitable for unexperienced users. A hanging
> > TUI without _any_ trace of reason, neither in the TUI no in the syslog
> > nor in the partman log must not happen.
> >
> > Raising the severity to serious because I think that this is a
> > scenario (reinstallation or retrying an incomplete installation) which
> > can happen often enough and will confuse users massively due no
> > feedback from the installer what happens.
> >
> > (D-I team: feel free to downgrade again to important if you disagree,
> > but I really would like to see this fixed for Jessie.)
>
> I don't think those usecases warrant to have this as a serious bug. That
> doesn't mean we shouldn't try and get this fixed (in wheezy and jessie).
>
> Any chance you could check and submit a patch?
I couldn't reproduce this issue with the upcoming D-I Jessie RC3:
Initial install:
- automated partitioning on a whole disc (/ + swap)
- proceed until the end
Reinstall with ext4:
- restart d-i
- manual partitioning
- switch the first partition (previous /) from unused to ext4, to be
formatted, mounted as /
- swap is automatically selected/used
- proceed until the end
Reinstall with btrfs (already reported as successful by Tianon on IRC,
cc'd to make sure):
- restart d-i
- manual partitioning
- switch the first partition (previous /) from unused to btrfs, to be
formatted, mounted as /
- swap is automatically selected/used
- proceed until the end
Reinstall with xfs:
- restart d-i
- manual partitioning
- switch the first partition (previous /) from unused to btrfs, to be
formatted, mounted as /
- swap is automatically selected/used
- proceed until the end
Reinstall with ext4:
- restart d-i
- manual partitioning
- switch the first partition (previous /) from unused to ext4, to be
formatted, mounted as /
- swap is automatically selected/used
- proceed until the end
So I think I'll stop trying to fix something I can't reproduce for D-I
Jessie RC 3 (https://lists.debian.org/debian-boot/2015/04/msg00264.html)
and move to other bug reports.
Mraw,
KiBi.
Attachment:
signature.asc
Description: Digital signature