The disk automount system has changed its behavior

[jrrpnani]

Let's see if anyone can give me a hand. I just made my user backup with BackinTime (an rsync manager) to an external hard drive formatted in NTFS. Before upgrading to Ubuntu 20.04 I have made my copies like this for months. The difference I have found is that now files with 'rare' characters for example 'Microsoft Corp.' (it finishes in point, a folder of fonts) and similar situations (folder Wine) now give me error and it cannot copy them. I smell that it has to do with as it reads the system the device NTFS and as it handles it
I have achieved that the folders created in my NTFS external hard disk manually mounted (I plug the disk to the system, it detects it, I unmount it, and I mount it by hand) do not give me that problem, so the issue is in how the system mounts them automatically. When I mount it manually I only indicate sudo mount /dev/sdc1 /media/user/disk without any option, I don't put anything special.
Any suggestions as to why you might have changed the behavior of the external disk after updating the system?

PD: I have it in NTFS to be able to take information to teams with guin2

[TheFu]

Back in Time cannot use NTFS storage to create proper hardlinks. It requires Linux file systems to work correctly.
Back in Time is really a hardlink manager more than an rsync manager. Hardlinks are are Unix capabilities. Further, POSIX permissions, users, groups, ACLs are part of backups. Generally, Linux access to NTFS storage doesn't support those things.

[jrrpnani]

But the strange behavior to which I make reference is that before the update to version 20.04, the sessions of BackinTime worked well so much with the same hard disk NTFS, like with Ubuntu 18.04. In fact the files that now give me error, I already have them in previous copies of the program.
The difference of mounted hard disk are:
mount by the system
'mount | grep /dev/sdc2' show:
/dev/sdc2 on /media/nani/Disco_Externo type fuseblk (rw,nosuid,nodev,relatime,user_id=0,group_id=0,def ault_permissions,allow_other,blksize=4096,uhelper= udisks2)
if I mount the disk:
/dev/sdc2 on /media/nani/DE type fuseblk (rw,relatime,user_id=0,group_id=0,allow_other,blks ize=4096)
Maybe this can help with the problem

[TheFu]

I've never use udisks/2. Mounting NTFS needs more control for my needs. I use autofs, which uses the automounter, with these options:

/misc/250G -nodev,permissions,windows_names,nosuid,noatime,asy nc,big_writes,timeout=2,fstype=ntfs,uid=1000,gid=1 000,fmask=0002,dmask=0002 LABEL=250G
In the autofs config file, only 1 line is allowed for each entry. autofs mounts storage upon request, which is good for USB and networked file systems. When no more files are "open" on the autofs mount, it will be umounted. This frees system resources when they aren't actually used.

For a manual mount, this would be the command:

sudo mount -o -t ntfs \
nodev,permissions,windows_names,nosuid,noatime,asy nc,big_writes,timeout=2,uid=1000,gid=1000,fmask=00 02,dmask=0002 \
LABEL=250G /misc/250G
I split the mount line into 3 lines using continuation characters to improve readability here. That's perfectly valid in any shell. Just be 100% certain there aren't any spaces after the '\' character or it won't work.

I'm mount to /misc/250G using the partition LABEL "250G" so I don't care about the device. A UUID for the partition could be used as well, but we humans remember labels easier. The options can be looked up in the mount.ntfs manpage, if you like. The big_writes and async are to improve performance. The windows_names option is to prevent any directories and files from having names that MS-Windows cannot support. The other options are for security reasons. My uid and gid are 1000 for each, which is just a coincidence, but extremely common. Check using the 'id' command. For NTFS, permissions can only be set during the mount. This is a huge failure when it comes to backups.

If you have any interest in using autofs, there is a guide here: https://help.ubuntu.com/community/Autofs
Just be aware that "automount" is a specific technical term that applies to using amd, autofs, automounter and not some GUI magical tool.

GUI programs often break standard stuff. Perhaps udisks2 has broken something here?

[jrrpnani]

I have solved the problem adding an entry to fstab:

/dev/sdc2 /media/nani/Disco_Externo ntfs-3g users,noauto,uid=1000,user,rw 0 0

It's all about user permissions that had change in any moment.

[TheFu]

I have solved the problem adding an entry to fstab:

/dev/sdc2 /media/nani/Disco_Externo ntfs-3g users,noauto,uid=1000,user,rw 0 0

It's all about user permissions that had change in any moment.

Comments about the "red" stuff:

/dev/sdc2 - devices change order from time to time. This is likely to break. The correct way is to use either the partition LABEL or UUID. The UUID won't change from boot to boot regardless of the device name. Uncover the current device name TO UUID mapping using the blkid command.

ntfs-3g is the F/LOSS implementation of the NTFS file system. For a long time, it is exactly the same as ntfs as far as fstab is concerned.

$ ll /usr/sbin/mount.ntfs*
lrwxrwxrwx 1 root root 13 Jun 20 11:58 /usr/sbin/mount.ntfs -> mount.ntfs-3g*
lrwxrwxrwx 1 root root 12 Jun 20 11:58 /usr/sbin/mount.ntfs-3g -> /bin/ntfs-3g*
They point to exactly the same tool.

Efforts are happening now by the company behind the proprietary NTFS driver to be included in the Linux kernel. They've submitted their code and it has been rejected twice with cleanup comments.
https://www.phoronix.com/scan.php?page=news_item&px=Paragon-NTFS3-V3-For-Linux Slightly interesting, but probably unimportant to us. The kernel reviewers have criticized it again, so the 4th attempt may be forthcoming.

You probably looked up the "users" and "user" options already. If "users" option is used, then there is little reason to also have "user" option. Having both is an opportunity for confusion and odd SW bugs to be encountered. Probably don't want to have uid= option specified either.

I always have to look those up in the mount manpage:



user Allow an ordinary user to mount the filesystem. The name of the
mounting user is written to the mtab file (or to the private
libmount file in /run/mount on systems without a regular mtab)
so that this same user can unmount the filesystem again. This
option implies the options noexec, nosuid, and nodev (unless
overridden by subsequent options, as in the option line
user,exec,dev,suid).

users Allow any user to mount and to unmount the filesystem, even when
some other ordinary user mounted it. This option implies the
options noexec, nosuid, and nodev (unless overridden by subse‐
quent options, as in the option line users,exec,dev,suid).

owner Allow an ordinary user to mount the filesystem if that user is
the owner of the device. This option implies the options nosuid
and nodev (unless overridden by subsequent options, as in the
option line owner,dev,suid).

With this information, I'd sudo mkdir /media/Disco_Externo and use:

UUID=xxxxxxxxx-yyyyyyy /media/Disco_Externo ntfs users,noauto,noatime,async,big_writes,windows_name s,rw 0 2
There are good reasons to avoid /media/{username} directories, mainly because other tools on desktop systems think they control those areas. If those tools get confused, bad things can happen.

Those numbers at the end of each line in the fstab matter too.

The sixth field (fs_passno).
This field is used by fsck(8) to determine the order in which
filesystem checks are done at boot time. The root filesystem
should be specified with a fs_passno of 1. Other filesystems
should have a fs_passno of 2. Filesystems within a drive will
be checked sequentially, but filesystems on different drives
will be checked at the same time to utilize parallelism avail‐
able in the hardware. Defaults to zero (don't fsck) if not
present.
2 is the correct answer.