I am running Ubuntu 16.04 server with several external USB HDD drives. After adding the most recent drive, it will no longer boot into ubuntu, but only into a limited cli. If I remove the entry from fstab and unplug the drive, then it will boot fine. I can then add it back to fstab and mount it and all is well.

Most of my drives are ntfs, but the drive in question is ext4. There is one other drive that is ext4 and it doesn't cause any problem.

Here are the entries I have in fstab

/dev/sdc1 /media/scott/FreeAgent ntfs-3g defaults,uid=1000,gid=1000,locale=en_US.utf8 0 0 
/dev/sdd1 /media/scott/My\040Book1 ntfs-3g defaults,uid=1000,gid=1000,locale=en_US.utf8 0 0 
/dev/sdf1 /media/scott/New\040My\040Book ext4 defaults 0 0 
/dev/sdb1 /media/scott/My\040Book ntfs-3g defaults,uid=1000,gid=1000,locale=en_US.utf8 0 0 
/dev/sde2 /media/scott/245E75E75E75B262 ntfs-3g defaults,uid=1000,gid=1000,locale=en_US.utf8 0 0 
/dev/sdg1 /media/scott/NextCloud ext4 defaults 0 0
  • Are you sure the mystery drive does not have a bootable partition on it? If it does, and if your BIOS looks at that drive first, then ... that's what you are going to get. – Peter Rowell Aug 15 '17 at 20:31

The situation you are describing might be related to different device enumeration orders. I.e. on some startups, the system HDD is another device file than on others. Certainity could possibly be achieved by seeing the full /etc/fstab and the exact messages being presented by the "limited CLI" (a screenshot before logging into that limited CLI would also do).

I suggest to replace the device-name-based entries in the /etc/fstab either by LABEL=... or UUID= s.t. all entries are independent on the order of which devices are recognized. To find out the UUIDs of devices, run lsblk as root.

Another thing to check is the BIOS: I know that my HP workstation does not boot if it has "boot from USB" as first entry in the BIOS boot order and there is any (even non-bootable) memory card in a card reader attached via USB. It is thus not sufficient that an attached device is "not bootable" for it to be ignored by the BIOS.

Finally, as long as there are "some errors about mounting", it is quite likely that boot fails eventough the mount fails on other devices. This is due to systemd's fail-fast behaviour with respect to /etc/fstab entries. An (insufficient, but quickly available) indicator for this is that once you exit the limited CLI (by just typing exit directly after login), usually continues the boot as if everything were OK. After startup (or from the limited CLI), check the output of the systemctl command for a first glance on the details.


It sounds like you have a bootable partition on this drive. If your boot order is messed up and you have a bootable partition on that drive, it will automatically boot onto that drive. If you are 100% certain that you have no bootable partitions on the drive, try running startx within the CLI that appears.

Report back if that doesn't work.

  • I double checked and it doesn't appear to be bootable, there is only one partition. There where some errors about mounting, but they where referencing other drives, not the problem one. I'll try the startx and report back – Scott Davis Aug 16 '17 at 21:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.