Mounting permanently a Storage unit in Fedora KDE (automount at boot, no password, all users can see and edit files)

It changes based on the SATA, Nmve or even IDE port used, right?
If I were to swap the 4tb and 2tb HDDs, they’d change their letter?

.

This was more a clarification I needed because of this:

I did not udnerstand what was said here.

.

It may just have been a mistaken read, or you believed that I made a mistake explaining myself.
I thank you for your concern, but not, i really just meant “work as in Windows”, as I described in more details in previous posts.

1 Like

Understandable, understandable, understandable, perfect.

A “place-holder”, a “virtual paperweight; same thing, different name (but I believe the one YOU used is the “professional one” tho).

At the moment I’d rather avoid working the 10tb drive on my main computer, and it’s already late here in Italy, so the work in the LivingRoom’s computer will have to wait a day or two.

.

Thank you people for your help here, if something will not work I will ask again in this page (since my goal has not yet been reached), but more than that you are not only helping me, you are also helping the other people who want to mount Storage Drives as I desire now.

.

I do not believe I’ll make a tutorial on the YouTube, but at least this page exists as guideline and clarification agent for those whom have the same knowledge as me.

here is what it means "mount the partition from drive sdc, partition 1 on mount-point /NEW-DRIVE, the type of file system is ext4, don’t fail to boot IF this disk is not connected, and the dump order is 0 and the dump level is also 0 (the dump options are archaic … do a man page on the fsdump command or google it :slight_smile: )

Isaac … "Oh Madonna … uno Italiano … " LOL!!!

.

So what you said is:

  1. Create a “place-holder” (“/NEW-DRIVE”, which I believe will change its “name” after I >>
  1. Create the New Partition (name: 10tb).

  2. Mount “10tb” into “/NEW-DRIVE”, so that I’ll have:

sdc
└─sdc1 "yadayadayada" part /10tb

Yes, that would change the letter. Beware, however, there are many other things that can change the letter, such as having a CD in your CDROM drive or an extra USB device plugged in that normally isn’t when you boot. It is recommended to use something besides /dev/sdc1 (or the like) in the first column of /etc/fstab. You can use a custom label or a computer-generated UUID (a big random number in hex that is difficult for humans to work with).


According to your earlier lsblk output the filesystem label on /dev/sdc1 is currently “10tb Test 01”, so you could replace /dev/sdc1 in /etc/fstab with LABEL="10tb Test 01". Of course, if you ever change that filesystem label, then the filesystem would no longer mount.

1 Like

Isaac,

Not quite right …

  1. The mount-point name does not change
  2. If you “Label” the drive with the name of “10tb” AND want to use it instead of the “Device Node” then the mount instruction in fstab would look like
    LABLE=10tb /NEW-DRIVE ext4 nofail 0 0
    and lsblk would look like
    sdc
    └─sdc1 “yadayadayada” part /NEW-DRIVE

Is this just a convoluted and confusing thing that someone coded in 30 years ago to have a laugh at us normal people?

Does this mean that “NEW-DRIVE” and “10tb” are now interchangeable names?
I wonder why this thing works in such an unintuitive way?

.

You said 1 “create new mounting point /New-Drive” and then 2 “mount the partition on it”.

Wouldn’t it make it

sdc
└─sdc1 “yadayadayada” part /NEW-DRIVE/10tb

??

Hi Isaac,

Is this just a convoluted and confusing thing that someone coded in 30 years ago to have a laugh at us normal people?

Yeah, it’s been this way for longer than that … LOL!!!
The reality of this is that the original intention was to give the system owner as much information and flexibilty as possible == there is a LOT of things you have to learn before all this makes sense sometimes :slight_smile:

Does this mean that “NEW-DRIVE” and “10tb” are now interchangeable names?

Only up to a point … the “/NEW-DRIVE” is “how to present the storage to the user” … the “10tb” is “what is to be used for this purpose” …

You said 1 “create new mounting point /New-Drive” and then 2 “mount the partition on it”.

Wouldn’t it make it

sdc
└─sdc1 “yadayadayada” part /NEW-DRIVE/10tb

look at the column headings in the lsblk output:

“NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS”

the “MOUNTPOINTS” refers to that directory we made in #1, /NEW-DRIVE

the /10tb" you appended to the MOUNTPOINTS, “/NEW-DRIVE/10tb” is not what happens. The “/10tb” is the name you gave to the disk, also known as a “Partition Label”. A disk label is just another way to refer to the device, /dev/sdc1. /dev/sdc1 and 10tb point to the same thing, a partition on a device.

I wonder why this thing works in such an unintuitive way?

LOL!!! … it may seam unintuitive at first … it’s one of those things that you need to do a few times to get used to it … just takes a bit of practical application :slight_smile:

Just a small correction here – partition labels and filesystem labels are not the same thing.

You can use lsblk -o NAME,PARTLABEL,LABEL to see both the partition label and filesystem label. In most places, what you will see listed is the filesystem label. You will probably need to use a partitioning tool to set the partition label (e.g. gdisk or parted). You can set the partition label if you want to, but then you would need to use PARTLABEL=<whatever> in /etc/fstab instead of LABEL=<whatever>.

Refering back to this

sdb is the 2tb SSD. Its 3 partitions are /boot/efi, /boot and /home.

.

What does

even do in this context???

Is it just “a slot/hole you dig in the OS’ memory to tell it there’s another drive” or does it do something “special” which changes something from THE QUOTED sdb SECTION ABOVE, or does it do nothing?

Sort of. It creates a mountpoint which is just an arbitrary directory name. You can use any name. I think einer made it upper case to indicate that it is variable — you can replace “NEW-DRIVE” with whatever text/name you prefer. It’s best to avoid spaces though (and / characters other than the first unless you understand what those do).

I have just now tried out what was advised here on the LivingRoom computer.

I have sudo mkdir /NEW-DRIVE-ONE, then with KDE Partition Manager I created “1tb” (simplified name).
I then edited the Mount Point to /NEW-DRIVE-ONE and then I did this:

sudo chmod -R 777 /NEW-DRIVE-ONE

Reboot.

.

.

.

Here’s the behavior I’ve witnessed on the computer:

1tb is mounted on /NEW-DRIVE-ONE.

My account (admin) is able to see the HDD mounted on Boot without password. It can create files which can be seen, used and modified by all accounts.

However, the guest_account could still see, use and modify admin’s files, but when it creates its files, these are not set by default as “modifiable (or usable, depending on the type) by all”.

By default one user is not allowed to modify the files that are created by another user.

sudo setfacl -R -m d:o::rw /NEW-DRIVE-ONE might be the simplest way to grant all users full access access to all the files in such a way that new files created in the future will also “inherit” full read and write permissions by all other users. That command will make a one-time modification of all files under the /NEW-DRIVE-ONE mountpoint.

Edit: Just a sec. I seem to have gotten that command wrong …


The following seems to work:

sudo setfacl -R -m d:o::rwx /NEW-DRIVE-ONE
sudo setfacl -R -m o::rwx /NEW-DRIVE-ONE
1 Like

Please, would you kindly either explain what

means (dissecting the parts of the commands), or even just point a wiki page where I can read it?

Wouldn’t running sudo chmod -R 777 /NEW-DRIVE-ONE on the Guest account too have the same effect on the files created by the Guest?

Yes, but you would have to run that every time you created new files. The chmod command’s effect is a one-time thing on existing files.

Inheritance of write permissions for other users is blocked by default for new files.

Sure. To break it down:

sudo = Super User Do
sefcacl = Set File Access Contol List
-R = Recursive
-m = Modify
d:o::rwx is a permission/access-conrol where:

  • : = A field separator
  • d = Default. This means what follows is to be applied to new files and directories when they are created under this directory
  • o = Other users
  • :: = An empty field. This would normally be for the name, but the “other users” type is a special case that matches all names. If the type were user or group, then corresponding username/groupname would go here.
  • rwx = Read, Write, and Execute. It is equivalent to one of the 7s in your chmod example.

For more details on such commands, you can read the manual (e.g., man setfacl).

HTH :slightly_smiling_face:

P.S. The reason for running setfacl twice is the first time is for setting the default/inheritable permissions, but the second time is to “retroactively” set the permissions on the files and directories that were already there before the default/inheritable permissions were modified.

1 Like

Gregory, Einer,
thank you for your extensive help both in time and action.

This is the closing message in this Post, meant to clarify what needs to be done to achieve [what I wanted in the Opening Post].
Today I’ve followed the instructions [as I will describe below] and the HDD automounts, at boot, without password, for all users, and [all its files, new and old, by any users, are seeable, accessible and openable by them all by default].

.

Einer’s instructions here tell you how to “create a new mount point” and "how to format your drive’s partition/s and how to mount them).

The “drive’s” name will not be its partitions’, but the Mount Point’s instead (it may be a problem for those who want to make more partitions out of the same drive, but I don’t and won’t bother with that) but the command which makes all files [“usable”] by all users does not seem to be the correct one.

.

For the correct command, Gregory got it.

Here he already explains why to use "his two commands instead than Einer’s for the last part.
The commands are:

.

.

.

.

With this, I feel confident in saying that “a solution has been found, and thus the issue has been resolved”.

A clarification about different partitions on the same drive may be added, but at least the basics have been achieved.

1 Like

Last thing, for specifications:

“The boot SSD” and “one drive, multiple partitions”.

.

If a second drive is added, like an HDD, and multiple partitions are built on it, do I have to make NEW MOUNT POINTs for each of them, or do I mount them all on the same point?

Therefore, if I have just and only a boot SSD on my machine (think a VERY cheap laptop, for practical use) can a partition be made out of that SSD, and then mounted on a NEW MOUNT POINT to allow all users to not only share the same set of desired files, but also to not waste space on the SSD by having multiples of the same file?

.

Separate question, which is more about computers in general to be frank:
I’ve read just and only opinions about this, I haven’t found a definitive source:
If an SSD is partitioned, is it just the Operating System (Windows, Linux, Mac, & others) which “are convinced” separate drives exist, but (as I read somewhere) "the actual SSD firmware still manages data among all its cells, maintaining the most longevity for the device,
or “an actual split” is made down to the cells level, making the partitions “age at a different rate” based on how much has been written on them?

Yes, you will need a new mount point for each partition. (Really it’s per-filesystem, not per-partition, but under normal circumstances partitions and filesystems are one-to-one.)

Yes.

It’s the former. Where the data is actually stored on the SSD is a bit of an illusion.

Here is a neat little animation from Optimizing Linux with cheap flash drives [LWN.net] :

segment

1 Like