Filesystem with last mount time in the 'future' halts boot with little help for user
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
util-linux (Ubuntu) |
Confirmed
|
Undecided
|
Unassigned |
Bug Description
Binary package hint: util-linux
I installed ubuntu 9.04 (since upgraded to 9.10) on my friends computer. Yesterday he called because his computer wouldn't boot since his brother had turned the computer off at the wall. I recognized this scenario from other experiences with some of my older computers, the hardware clock had been reset because the cmos battery was dead.
When I was checking his system I noted that the error message wasn't particularly helpful, from what I can recall (not sure if I can reproduce in a VM) the 'last mounted in future' part was not visible. Changing the grub entry by removing 'splash' allowed me to see the full error message and confirm my suspicions. I then went on to run fsck -y /dev/sdxx (ext3) which repaired the problem and allowed the system to boot.
I have encountered this scenario on a number of occasions and while I am able to recognize the problem and implement the fix, 'average users' may not, particularly considering that the best indication of the specific error is hidden due to the splash. I think this could be improved by either of the following:
* Ensure that the relevant error message, and perhaps simple instructions for repair are visible to the user when the error occurs. This may mean asking them to run fsck and reboot or check there hardware clock, both of which may be to technical a task for 'the average user'
* Modify the behavior of the system such that this error is automatically repaired rather than forcing the user to repair the problem. The quote below is from a debian forums post here: http://
My only concern with this is whether the clock will be set correctly by the system once booted, is ntp support installed by default now?
QUOTE:
"
/etc/
[problems]
# Superblock last mount time is in the future (PR_0_FUTURE_
0x000031 = {
preen_ok = true
}
# Superblock last write time is in the future (PR_0_FUTURE_
0x000032 = {
preen_ok = true
}
This config then tells e2fsck that the superblock mount and write time issues are ok to be auto-fixed (preen_ok=true) and that I don't want to see e2fsck mentioning it (preen_
The numeric codes (0x000031, 0x000032) are not really in the man page, but you can find them in e2fsck sources (http://
I'm having a similar problem on O2 Joggler hardware which does not have a hardware clock, and so the clock is ALWAYS wrong until after the network comes up and NTP can work.
Ubuntu Netbook Remix Lucid Beta 2 goes in an inifinite loop from mountall of trying to fsck the driver but failing with exit code 4.
I can edit /etc/init/ mountall. conf to make it fsck first (hard wired) by why - there is nothing wrong with the disk, I don't want all the delay of fsck, I just want the e2fsck tools to realise that the superblock time being in the future doesn't always indicate something about the file system but often something about the hwclock, or lack of.