Better to redo entire server or just delete pool?

I set up a backup server a couple years ago with Scale. Worked great. However, after moving, and upgrading one of my computers to Windows 11, all the subfolders have become read only. I can only write to the main drive. I’ve tried all the fixes I’ve found mentioned and nothing seems to work. I am currently backing up the server before messing with it any further. (Most was already backed up, but better to be safe)

So, options are, delete the pool and set up a new pool without messing with the base TrueNAS or just starting over and reinstalling TrueNAS.

The reinstall option might be best because I have not been able to update, partially because I set it up not to connect to the internet, but just a local LAN, though everyone on the WLAN can see it. I am on version 22.02.4.

I am thinking reinstall the pool might be faster. I am concerned though, that the same problem will happen with the new pool, as it seems to be a permissions issue?

Anyway, Thoughts? Other ideas? I’m willing to try anything at this point.
Thanks!

Yeah this definitely seems to be a permissions issue. Can you give more information?

Are you using the same user to access the share that you used before?

Are there any differences in the permissions between the top folder and the subfolders?

I would say that you could go into the top folder permissions and apply the permissions recursively and to the child datasets. This could cause other issues if you do have them set up differently on purpose. I would try and look at the differences in permissions before doing this. It is kind of like a hammer vs scalpel.

2 Likes

Agreed. A recursive change of permissions will run a lot faster than rebuilding a pool from scratch.

1 Like

Have you tried setting up a new child dataset for SMB sharing and following the TrueNAS Documents? If the new, test share works, we should be able to get your other working also.

I keep it simple. Root dataset then a child dataset set up just for sharing SMB. That dataset is owned by ‘smbuser’. That user is set up just for sharing.

Browse some other threads and do the Tutorial by the Bot to get your forum trust level up to post images. We will want to see how you have your datasets and SMB set.

TrueNAS-Bot
Type this in a new reply and send to bring up the tutorial, if you haven’t done it already.

@TrueNAS-Bot start tutorial

Maybe I have too many users. I initially had trouble logging in as the same user on different computers, so I set up a user for each computer, if that makes sense. But then I think I figured out how to used the same user. And maybe that is why it suddenly isn’t working?

I may have set it up all wrong also. I set up a pool that shows on the storage. It doesn’t show the child dataset. But the child dataset shows on the shares? Also, Is there another place to change permissions other than in the credentials area? In the root user under “other” I allow write along with read and execute, But on the root dataset, other only allows read and execute.

I am sure it is a permission problem, but I don’t know how to fix it. I have everything I can find changed to read/write/execute.

Thanks

This is interesting. My share dataset does not show up where the new dataset is. And, the YouTube video I am following has a section called Datasets, whereas my version does not. I set it up under the Storage section. The dots by the pool gives the option of creating a dataset. I actually can only find the dataset (folder) I use in the SMB section of shares. I do think I set it up incorrectly. It worked until I moved. I took apart the server to transport it.

Anyway, set up a new dataset per the instructions and everything seems to be working like they are supposed to. I think what I will do is move the data to the new backup dataset and delete the old backup.

And then I screwed it up. why do I do these things. I tried matching the permissions. Yeah. Seems to have crashed the server. The problem with having more than one computer to work on at the same time.

If it comes back, I will finish the backup. Then try to figure it out.

What video is that? Most YouTube videos on TrueNAS are garbage (to be polite) or even dangerous. Tom Lawrence is the notable exception here and probably a few others. Avoid LTT by any means necessary

I’ve watched quite a few LTT videos where TrueNAS was a topic but I can’t remember them ever going deep into any TrueNAS configuration…
Do you have an example of them producing TN-related garbage?

Linus has admitted repeatedly how he’s made mistakes configuring TrueNAS systems in the past which is why the collab between iXsystems and LTT (among other mentions) caused A LOT of head scratching in the old forum.

Some of the examples you’re looking for (like Linus drilling holes into motherboards and being surprised they don’t work afterwards anymore) are mentioned there giving you the keywords you need to investigate further.

Bottom line, Linus may have been an excellent self-promoter of perhaps amusing content but Lawrence Systems, @yorick, @Stux, etc. are among few content creators who I’d follow on YouTube re: anything TrueNAS related.

3 Likes

A video by a guy at Lawrence Systems? Worked fine. I can read/write/execte with it. Owner is root. The key seemed to be setting it up as an smb and not a generic dataset? The ACL for my original backup dataset shows as Posix-ie? The new one doesn’t have that. The old one requires a mask user also. The new dataset doesn’t. I’m really thinking I set the original up incorrectly.

1 Like

After going through all of this, I definitely set it up incorrectly. I have a Pool. I have a Dataset. They do not appear to be related. Much as I hate to, I think I need to start from scratch. The data is all backed up. Most in multiple places. I currently can’t move anymore data from the old dataset to the new one because the space is not freeing up as I delete folders, even after days and multiple restarts. I believe it is a snapshot, but since the dataset doesn’t exist, it is not showing up anywhere that I can see.

So question - I set up a pool. Then I set up a dataset. Now do I set up a child dataset? Or is the first dataset fine?

This is primarily a back up of my music production programs, a number of which I can no longer re-download or it will cost money to redownload. The virtual instruments are what takes up the majority of the room. But I also like having everything all in one place when I set up a new machine.

Thanks so much for everyone’s help.