Hey everyone..

I have an usb 2.0 drive that I used YaST to format with EXT3 (Because of File Size Limitations).
If I automatically mount the drive on boot, I get a crash saying that my drive failed fsck.

It appears that the USB drive doesn't wake up in time for it to be mounted. However, I didn't seem to get this problem when out of the box the drive was formatted with fat.

Has Anyone had a similar problem? Anyone know how to fix it?

I would prefer not to have to mount the drive manually since I'd like to store automatic backups to the drive using a scheduled job.

My fstab

/dev/disk/by-id/ata-HDS722516VLSA80_VN6D3ECDDXE6LD-part1 swap swap defaults 0 0
/dev/disk/by-id/ata-HDS722516VLSA80_VN6D3ECDDXE6LD-part2 / ext3 acl,user_xattr 1 1
/dev/disk/by-id/ata-HDS722516VLSA80_VN6D3ECDDXE6LD-part3 /home ext3 acl,user_xattr 1 2
/dev/disk/by-id/ata-ST3500630AS_6QG16L7W-part1 /windows/C ntfs-3g users,gid=users,fmask=133,dmask=022,locale=en_US.UTF-8 0 0
proc /proc proc defaults 0 0
sysfs /sys sysfs noauto 0 0
debugfs /sys/kernel/debug debugfs noauto 0 0
usbfs /proc/bus/usb usbfs noauto 0 0
devpts /dev/pts devpts mode=0620,gid=5 0 0
/dev/disk/by-id/usb-WD_10EACS_External_57442D574341553437373433393533-0:0-part1 /media/ExternalBackup ext3 defaults 1 2

Is it a Seagate external drive? The newer Seagate external drives are known to go to sleep and spin down. Linux doesn't wait long enough for them to spin back up.

If it is a Seagate drive then this has been addressed in other posts. Use the LQ search to find those posts or use Google. You should always search for the answer to a question before posting it. Most questions have been asked and answered numerous times.

Thanks stress_junkie

However, the drive in question is a Western Digital "dev/disk/by-id/usb-WD_10EACS_External_57442D574341553437373433393533-0:0-part1 /"

I have searched quite a bit actually...

I am assuming that your external disk is OK, and passes a manual fsck, but something is wrong with the way it is being spun up or detected.

If you remove the 2 at the very end of the line in /etc/fstab that refers to your external disk, it should not be fsck'd

Also, see the -i option to tune2fs

If you do this, you should remember to run fsck manually on it from time to time.

I'm having the same problem with a Maxtor OneTouch 750G (USB). Intel Mac Mini running Ubuntu server 9.0.4 (2.6.28-14-server).

If I disable fsck, the system boots but the drive doesn't even get mounted. I need to have the drive properly fscked and mounted during boot to ensure it is ready to go before certain userland daemons are launched.

I tried the disabling STANDBY and although the drive accepts the change, it does not address the problem. The drive is already spun up well before fsck is attempted. I think the system is just not allowing enough time for USB drives to be recognized.

I suspect that just adding a 5 second delay at the right point in the boot sequence should do the trick. Trying to figure out how to do that...

You could write yourself a little script:

If the device isn't plugged in exit 1
If the device is mounted, unmount it
Run fsck on it, if that fails, exit 2
otherwise mount the device
If the mount failed exit 3
otherwise exit 0

If no errors, then start your "userland daemons".

Welcome to LQ.

Thanks, I think I'll probably just do that. But I think there is a general issue that the kernel needs a parameter to just give more time for USB drives to come online. If it's not a known issue, maybe I'll take a closer look under the hood.

seanadams -

Have you been able to discover any other solutions? It had occupied enough of my time attempting to fix it, but I have some time so I'm going to try and revisit it.

Thanks

No, not for USB, but the same drive works OK on firewire. I remain convinced that this is a general bug in the kernel.

But... probably offtopic, as it happens, I have abandoned this computer because it seems to have other issues. If I use "stress" to simultaneously pound the CPU and internal HDD it eventually crashes in a few hours. I just got another unit to replace it so I'm about to find out if it was one-off bad hardware or not.