*** This bug is a duplicate of bug 65683 ***
https://bugs.launchpad.net/bugs/65683
Even though I'm hearing that this isn't a bug, but moreover a feature
that was intentionally removed, it's still an issue in Jaunty, and this
functionality never should have been removed. Also, please note that
this needs to work both in the splash screen AND when quiet and nosplash
are not being used.
This "bug" bites me all the time. Autofsck has another problem that's
even worse, though, so I had to remove it: If Autofsck prompts that a
disk needs checking, and I select "OK" on reboot, the system comes back
up, checks the disks, then shuts down!! I'd rather sit through the disk
checking with no warnings when I reboot than have the computer shut
down. Really, though, cancelling a running fsck should be included in
ALL linux distributions and having a tiny piece of software like
Autofsck that has its own problems is not a solution.
I'm running Jaunty, and I boot verbose (I removed "quiet" and "nosplash"
from my menu.lst) for other reasons. Pressing ESC, ctrl-C, ctrl-D, ctrl-
alt-delete do nothing. (well actually ctrl-alt-del does cancel the fsck,
but you have to press ctrl-d IMMEDIATELY afterwards (and I mean
literally immediately) to avoid a reboot (which just triggers another
round of fsck's)... and even if you DO manage to press that quick
enough, there are still several daemons that don't start, like
networking for example, that must be started manually). no networking,
no ssh, no remote management, no server useability. This is quite a
conundrum that could easily be solved by not disabling the ability to
cancel and not implementing such a policy in future releases.
My /, /boot, and /home are on smaller high-availability RAID arrays, and
not only that, my / is ReiserFS so is barely affected by this problem.
Checking just those wouldn't take very long. The problem is I have
nearly an additional 1.5TB of ext3 data across several drives in
different RAID configurations. It's these checks that really kill me,
and they could be all run while the system is fully up and running in
runlevel 5. I don't want to change the frequency of checks, either,
because if I don't NEED to use the box, I can just let it do its
scheduled fsck and when it's finished, it's finished.
What would really be nice is if the functionality that should be there
(ctrl-c to cancel on an ad-hoc basis) was still there. The decision to
remove this feature was not fully thought out, and it should be there
once again, as it always has been in Debian, RHEL, and used to exist in
Ubuntu until somebody made the unwise decision to disable it.
Again, another piece of software, i.e. autofsck is NOT a valid solution.
--
forced disk check after every 20 mounts should be postponable
https://bugs.launchpad.net/bugs/222112
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
ubuntu-bugs mailing list
ubuntu-bugs@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
