General :: Boot Up Misbehavior After An Update During Fsck
Jun 25, 2011
Today I did an update on my Debian 6.0.0 installation, which included a kernel update. I also did an Nvidia driver update, afterward. All well, until upon a third boot up an fsck check was forced (34 mounts). Some odd behavior ensued. At about 19% it exited the check, printed what appeared to be a load of inode numbers with error messages, and "froze" with the caps lock and num lock lights on the keyboard flashing.I did a couple of grub recovery modes, each which advanced a little further, and all is well again. However, it concerned me. Where in var/log do I find the file of that particular fsck activity?
I wanted to know how can I make fsck automatically fix HDD errors when it needs to check HDD during boot up and do not ask user for it . I found some of fsck related scripts in /etc/init/ and /etc/init.d/ but I don't know what to do with them.
My computer got hung up restarting after an update yesterday so I switched it off and back on, which led to the error message "No init found. Try passing init= bootrag", looking that up I found a thread that recommended running fsck to fix the errors.
I first went to disk utility and did the check filesystem and repair button but that just had a pop-up that said "File system is NOT clean." So I went and tried fsck on the drive which gave me the error message "fsck.ext4: Device or resource busy while trying to open /dev/sda1 Filesystem mounted or opened exclusively by another program?"
Figuring this may have been caused by the disk utility I restarted and tried fsck again without opening any other program, and I still get the same error message. What is causing this? Swap? What can I do to get fsck running to fix the original problem?
I have my drives set to run fsck on boot up. This is the default setting for Fedora but if there is a problem it prompts me to run fsck manually. Is there a way to have Fedora just run fsck and fix any errors it finds on boot up?w in Debian distros you can put FSCKFIX=yes in the /etc/default/rcS file to do this. But I can't find the equivalent on Fedora.
I am wondering if there is a way to manually trigger a file system check during boot by pressing/holding a key. Maybe there is already a keyboard shortcut built in to do this?I know that using tune2fs you can modify the number of boots (mounts) between file system checks, or even use "shutdown -rF" to reboot the system and force a file system checkAlso, I do not want to force the user to choose to run/skip the file system check during boot. For example, prompting the user with, "Do you want to run a file system check [y/n]?" each boot (or even each time the system thinks it should run a file system check),s not desirable
I need to boot in WXP just because the guys at the company don't want to update for Linux drivers in the TS so I can't connect from home unless I do it in Windows XP. But today WXP refuses to boot, when the progress bar appears, the system boots automatically. The WXP CD doesn't help, the repair console doesn't let me run chkdsk or anything (dir C: gives an error about listing addresses or something like that). Is there any way to use fsck or anything in Linux to repair the WXP NTFS partition?
i have the same issue but with a real install, not a vmware.every reboot fsck says that my main partition was not cleanly unmounted and check is forced.during shutdown the ubuntu does not show messages in a clear way but in random places. however, i am able to see the "/ is busy" message 1 second before reboot/off.my distribution is 10.10 upgraded from 9.10, 10.04
I was editing my fsck at etc/fstab/ after removing a hard drive and i think i commented out the wrong drive . now i just get a blank screen on boot. I have the ability to use my ubuntu cd for a recovery console, live desktop, etc.
UPDATE: This is a bug: [URL] Evidently the problem is with plymouth because a workaround is to add "rd_NO_PLYMOUTH" to the kernel boot options. I don't get a prompt for my disk encyrption pass phrase---just a flashing cursor---but that's a small price to pay for being able to run fsck when the root filesystem wasn't umounted properly.
I have fully updated f13 (as of today) on a laptop with all ext2 file systems (It has nothing but flash memory.) If it's shut down without unmounting all file systems, it drops to a shell and asks for the root password to run fsck when it's rebooted. Every key press is treated as though it were <enter>, with a response to the effect that the password is incorrect.
I just installed 10.04 from the minimal install cd (twice), and when I boot it up, it runs fsck on the 3 partitions, shows they are clean, and then nothing. I can't even get to recovery mode. I had previously installed 9.10 fine on this computer. I can't even find out what the real problem is, even with the quiet splash disabled. It just stops after successfully running fsck.
Today for strange reason, one of my Ubuntu 64 bit server Linux VM failed to start ? it stopped in the FSCK scan status with the error as attached. few days back I've added new hard disk successfully and format it as sdb5 without problem.
I have a 2TB file-system and when the machine reboots it fails the fsck, halts and goes into maintenance mode.Stats: I have have RHEL 5, 2.6.18 kernel, the file-system is an ext3. The file-system is on an EMC AX4 connected with fiber channel HBA.So far my reading tells me this should work because under 2.6 4TB is OK. Any ideas why this fails?If I take it out of the fstab file and mount it manually the boot is OK and the file-system behaves well. I can change the fsck check option in the fstab to 0 but I don't think I should have too. Everything I read says that 2TB ext3 file-systems are OK.
I was running 11.0 and it stalled in the middle of a number of updates. On reboot I now get (if I remember correctly) Error 15: File not found.I managed to get my hands on a 11.0 live cd and started trying to fix grub. However, when I try mounting the harddrive using "mount /dev/sda1" I got
Code: can`t find sda1 in /etc/fstab or /etc/mtab "mount /dev/sda1 /mnt" gives
I need to be able to login to terminal and I can't seem to find an easy way to do it from the login screen, or elsewhere that I would normally find in an older version of ubuntu. I need to run fsck with drives unmounted
all is well on my headless Lucid server until a recent apt-get upgrade && shutdown -R now ... it did not come back up? after i moved a screen to the other side of the house, i found fcsk waiting for input during the boot process errors on / ... (I)gnore / (F)ix " ...so i had to attach a keyboard just to push <F> i could change /etc/fstab so it never runs fsck, but this doesn't seem wise. how can i make it <F>ix automatically ? ( or maybe after Xsec )
I have ran into a bit of a problem. my server(10.10) is hosting a website with no problems whatsoever. I decided to install an extra hard drive, and format. i did both, and turned off my server after some more configuration. After a while, I decided that the new hard drive was a little loud for my liking, so I removed it. I had not put any files on it. I rebooted and saw a message that said this:Quote:fsck from util-linux-ng 2.17.2/dev/sda1: clean, 101202/2351104 files, 1516082/9393920 blocksI think that when I turned it off, it was formatting. How do I stop this from halting the boot? I can get by it by pressing the b or a key.
I have $ uname -a Linux kub 2.6.32-5-amd64 #1 SMP Mon Mar 7 21:35:22 UTC 2011 x86_64 GNU/Linux
Most of the time when I boot my PC I get an error about fsck.ext4: Unable to resolve... I don't know why it's happening.
The problem is happening with my external drive that has 3 partitions: /dev/sdc1 /dev/sdc3 /dev/sdc2
About 90% of the time I boot I do get the error. Sometimes after getting the error I can login and the external drive (/dev/sdc) is already mounted: $ df -H Filesystem Size Used Avail Use% Mounted on /dev/sda2 15G 8.0G 5.8G 58% / tmpfs 1.9G 0 1.9G 0% /lib/init/rw udev 1.9G 246k 1.9G 1% /dev tmpfs 1.9G 738k 1.9G 1% /dev/shm code....
The UUID's in the error file match the output of the command blkid. And the UID's of blkid match the fstab UUID's. I don't know what to do at this point.