I'm using a RPi B+ with XBian and transmission, and an external USB 2TB disk, which has an external power supply.

That disk is NTFS, and I've noticed that the CPU usage of the mount-ntfs process is really high..

What can I do to improve this situation, short of using ext4 that would impede the usage of that disk on other non-Linux systems?

2 Answers 2


There's a kernel driver; Xbian's mount-ntfs is presumably a user space one, but the former is a different implementation. I would guess it will consume less resources when you are not actually doing anything with the partition.

However, it may not be available on Xbian. Unmount the drive and try instead:

mount -i -t ntfs /path/to/partition /mount/point

This should avoid invoking the userspace driver because of -i ("-i, --internal-only: Don't call the [userspace] helper even if it exists <- man mount). If you don't get any errors, check ps -A | grep mount. There should not be any such process running; it will exit right away having invoked the kernel driver if present.

If there is such a process, it's probably the same thing as mount-ntfs. If not, the partition should be mounted and available.

  • Done, there is still the mount.ntfs FUSE process running, so no result there :( Jan 12, 2015 at 21:43
  • Don't know why, maybe thanks to a second reboot, now the root process kicked in :D I hope that helps, thanks! Jan 15, 2015 at 21:57
  • New update: had to remove, NTFS-3G brought instability to transmission... Jan 16, 2015 at 18:43
  • You also might want to add -i option to mount to ensure kernel space driver is used and not fuse. Although IIRC it has some limitations.
    – mlt
    Oct 29, 2015 at 16:54

Add -o big_writes to your mount commmand or big_writes to /etc/fstab

So the mount command above would become:

mount -i -t ntfs -o big_writes /path/to/partition /mount/point

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.