Newbie - Unassigned Disks? + SMB set up

TrueNAS General
, , ,
1

Hello,
I’ve set up my first Scale 24.04 home NAS last weekend. I continue to read the forums and documentation but I’m not finding answers to my questions. I suspect that I don’t know what to search for.

I installed 23.10 last weekend to a new SSD and updated to 24.04 this week. Everything went smoothly. Bravo to the folks who built and tested the release and upgrade process!

After the fresh install I imported my ZFS drives from a previous nas4free/XigmaNAS NAS. There are home folders and media share on the ZFS drives (mirrored 8GB drives)

  1. Today I see a note “2 Unassigned Disks”. These disks are already in the pool that is active on the TrueNAS. Why would there be 2 Unassigned Disks? Should I be adding these drives to the same Pool1 again??

  2. First issue is that I can’t get SMB to work properly.
    2.1) Where can I do some reading or see some examples of correct SMB set ups? (I haven’t been able to Google anything so far. Uncle Fester has been helpful but I’m still missing something on how this is supposed to be set up. I was able to do this on the XigmaNAS so I’m not sure what can be different here.) I think I’ve messed up the ACLs because I have an error “can’t read superblock” when I try to connect.
    2.2) I created a number of users but no user can write any files to the share folder despite having RWX privileges. These are all media files. I can mount the share, read the files but can’t copy anything to the share.
    2.3) the home folder for the user never comes up when connecting via SMB. I can see the /mnt/Pool1/home/user1 and files within it when I connect via SSH but connecting via SMB never goes to the home

Additional background info that might or might not be relevant.
Initially, the ZFS import has a problem. I ended up with /mnt/mnt/Pool1. The old forums helped me to fix this. So now I have /mnt/Pool1. I have both /mnt/Pool1/Home/ and /mnt/Pool1/MediaShare. I’m not sure if this could have screwed up something for SMB.

Thanks in advance.

2

I would suggest that you focus first on the ZFS issue - because you may not be able to tell whether the SMB symptoms are a consequence of the ZFS or SMB issue.

To help with this we will need more details of what the zfs status is and how it differs from what you are expecting.

In the mean time, I would try not to do anything on the zfs pool just in case it makes any corruption worse.

3

Thank you @Protopia That make total sense.

The zfs status looks exactly as I expect.

When I imported the pool from the old NAS, it went ok except for attaching to /mnt/mnt/Pool1. This was fixed using the info from a few Forum posts. I also received the notice to update ZFS since the old NAS used an order version of ZFS. I did the update via the TrueNas GUI. There were no errors.

Here is the output of zpool status

sudo zpool status
pool: Pool1
state: ONLINE
scan: resilvered 3.35T in 05:45:49 with 0 errors on Wed Aug 5 22:04:15 2020
config:
NAME STATE READ WRITE CKSUM
Pool1 ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
ata-ST8000VN004-2M2101_WKD2DRKJ ONLINE 0 0 0
ata-ST8000VN004-2M2101_WKD2JHNN ONLINE 0 0 0
errors: No known data errors
pool: boot-pool
state: ONLINE
scan: scrub repaired 0B in 00:00:23 with 0 errors on Sun Apr 28 03:45:25 2024
config:
NAME STATE READ WRITE CKSUM
boot-pool ONLINE 0 0 0
sdc3 ONLINE 0 0 0
errors: No known data errors

The weird thing is this new info in the Storage Status page.
I don’t have any unassigned disks. There are two disks in the Pool1 mirror and a boot SSD.

image
image1200×890 32 KB

4

I think I have fixed the SMB issues but I still have this “Unassigned Disks” issue.

The 2 disks (original ZFS import from other NAS) that are Pool1 and operating fine except sometime in the first week they started also showing up in the “Unassigned Disk” list. I am concerned that if I add these disks to the Pool1, it will overwrite or screw up a properly operating Pool.

Would someone have a suggestion on how to go about figuring out what has happened and how to fix it?

Thanks again for any help.

5

Compare these outputs, to confirm that you are indeed seeing unassigned disks.

lsblk -o name,model,serial,partuuid,fstype,size

zpool status
6

Thanks @winnielinnie

From the lsblk -o name,model,serial,partuuid,fstype,size
I get the following. If I compare the boot SSD “sde” with the ZFS mirror “sdb” and “sdc” it doesn’t look right.
lsblk -o name,model,serial,partuuid,fstype,size


NAME     MODEL                SERIAL          PARTUUID                             FSTYPE       SIZE
sda      WDC WD40EFRX-68N32N0 WD-WCC7K1ZU2FYJ                                                   3.6T
└─sda1                                        f4eef083-9426-423d-81e5-5fc32ad502f1 zfs_member   3.6T
sdb      ST8000VN004-2M2101   WKD2DRKJ                                             zfs_member   7.3T
sdc      ST8000VN004-2M2101   WKD2JHNN                                             zfs_member   7.3T
sdd      WDC WD40EFRX-68N32N0 WD-WCC7K7LYUYEY                                                   3.6T
└─sdd1                                        cada6eca-af66-45a8-9398-f04446cb2fbe zfs_member   3.6T
sde      CT500MX500SSD1       2350E8896868                                                    465.8G
├─sde1                                        eb5cbf99-5ca9-4340-a9e9-4309e0e85e12                1M
├─sde2                                        554651e6-1de5-4c46-91e7-00fb030337eb vfat         512M
├─sde3                                        97db2c7a-1aa3-473b-b135-20038989176b zfs_member 449.3G
└─sde4                                        fa63c387-a63a-416a-a2e7-fe6b3cea333e               16G
  └─sde4                                                                           swap          16G

I did add two more disks this evening. I’ve left those out of the list as they were not imported but rather I created a new second pool. So the numbering of the disks has changed.
From zfs status I get the following:

pool: Pool1
state: ONLINE
scan: resilvered 3.35T in 05:45:49 with 0 errors on Wed Aug 5 22:04:15 2020
config:

NAME STATE READ WRITE CKSUM
Pool1 ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
ata-ST8000VN004-2M2101_WKD2DRKJ ONLINE 0 0 0
ata-ST8000VN004-2M2101_WKD2JHNN ONLINE 0 0 0

errors: No known data errors

pool: boot-pool
state: ONLINE
scan: scrub repaired 0B in 00:00:23 with 0 errors on Sun Apr 28 03:45:25 2024
config:

NAME STATE READ WRITE CKSUM
boot-pool ONLINE 0 0 0
sde3 ONLINE 0 0 0

errors: No known data errors

pool: Pool2
state: ONLINE
config:

NAME STATE READ WRITE CKSUM
Pool2 ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
f4eef083-9426-423d-81e5-5fc32ad502f1 ONLINE 0 0 0
cada6eca-af66-45a8-9398-f04446cb2fbe ONLINE 0 0 0

errors: No known data errors

It looks like something is 100% right but I’m not sure what it should look like. :slightly_frowning_face:

Edited to add back new disks installed this evening.

7

The output is incomplete, then?

I notice your two 8-TiB drives lack swap partitions. (Actually, partitions, period.)

Normally, the disk(s) are partitioned, and it is these partitions that are used to build your pool’s vdevs. (Yours were not partitioned.) This also explains why they lack a PARTUUID. Perhaps this is throwing off the GUI / middleware?

EDIT: It makes it more difficult if things are intentionally omitted. Even your screenshot doesn’t show any information aside from “2 unassigned disks”.

8

Thanks for taking time to review my post @winnielinnie .

I understand but I thought that I didn’t want to further confuse the issue by adding new disks that weren’t related to the issue at hand. I’ve added that info back and hope it is helpful.

I’m not sure if it is good practice to do the add back of info but hope that someone will let me know as I want to be a good member in this community.

As to the screenshot, it was only to show that this message is popping up in the Storage tab. I’m not sure how much more I can add without doing multiple screenshots. In any case, I’m happy to provide any info in order to help understand what’s going on and how to resolve it.

I added back the output that I edited on the two disks that I added last night. These were used to create a new Pool2.

9

If you click “Manage Devices” under Pool1, what does it show?

If you click “Add To Pool” under Unassigned Disks, what does it show?

What are the full outputs of:

lsblk -o name,model,serial,partuuid,fstype,size

zpool status

Do you have swap creation configured for the data disks? (Is it set to any size in GiB besides 0?)

10

Here is what it shows:

image
image1566×798 19.7 KB

image
image1548×688 38.3 KB

I added the full output of the two commands back into the reply above. Sorry if that causes problems. I’m feeling a bit overwhelmed at the moment.

The Pool1 was created as a default ZFS data storage disk for my old NAS which I’ve now replaced with TrueNAS. I don’t remember how it was created or if the option was even there to create a swap. It was about 10 years ago and ran without issues. It sounds like what I did 10 yrs ago was stupid …

11

Is that really what SCALE shows? Just “7.28 TiB HDD x 2”? It doesn’t even give you the actual disk names / IDs? :neutral_face:

My hunch is that since Pool1 was created without partitioning, it is throwing off the TrueNAS GUI / middleware into believing that those two particular disks are not assigned to any pool. (Using strictly the command-line, everything looks correct in terms of disks, vdevs, and pools.)

You’ll notice with Pool2, it lists the member devices by their PARTUUID, instead of model + serial number.

@kris or @Captain_Morgan is this the expected behavior?

12

Yes, I can confirm that is the case.

13

Should I be filing a bug report for this?

1 Like
14

I see that @mooglestiltzkin is currently typing up a reply.

Brace yourself for a flood of text that will make a PhD dissertation blush… :umbrella:

torrent-from-above

3 Likes
15

i’m new to truenas myself but this was what i learnt.

for my media datasets, when i create the dataset, i set it to smb from the default generic. i have a couple of media datasets similarly setup as such. 2 of which i select advance, enable encryption, passphrase. the other datasets i did not use encryption. For these media datasets, you want to create a new user, and assign it to a group of your choice, or simply pick the existing inbuilt administrator or inbuilt user. To update your acl changes, go to truenas services, disable smb, then re-enable. at this point the smb should work. If you are on windows go to network, browse to your truenas to open, it prompts for credentials, enter them then it ought to work. You can test read/write. It’s dependent on your ACLs.

i also use jailmaker for docker jail. for that separate dataset, i left it at the default generic. In the jailmaker doc, they said set this as such then don’t make changes by adding acls and stuff via truenas.

there is another dataset for docker, this i also set to generic. i also did not mess around with acls for this.

I did not setup nfs, though some people may want to do that. but i won’t get into that since i didn’t set that up.

afaik jailmaker does not have acls setup for it, and i didn’t add any smb service for it so how exactly do i make changes and stuff? Well i assume you login via shell then do it all commandline. If you are wanting to do stuff for docker jail, go to truenas shell , jlmkr start docker, then jlmkr shell docker. once inside you can cd to change directory, use the bindmounts set in jailmaker.

the alternative method i used that is probably not recommended, i enabled ftp service, then allow root to connect to it. Then using winscp i login as root then am able to browse around through docker and jailmaker datasets that way via sftp. Maybe there is a better solution for this that is less of a ops sec risk :sweat_smile:

These are some of the resources i used to learn setting up ACLs and permissions for datasets

I’m also still learning myself. but i find this stable as of now for my specific simple basic usage.

2.2

Hm… not sure whats going on here. But the data i recovered from my backup to my truenas, it had rsync set to set permissions to destination. So whatever i had set when i created the dataset (on the truenas), would also be applied to the incoming data being restored from the backup. For myself i didn’t encounter any issues with permissions on that end. Maybe you did something in acl that borked things :grimacing:

2 Likes
16

I would. The middleware / GUI needs to handle the underlying devices with more grace. It’s not a good sign when you have two contradictory claims between the middleware and the command-line.

17

Thanks @mooglestiltzkin . Wow! It will take me a bit to absorb all your help. Thanks!! I’ve already watched the first and last video. Both are helpful.

Thanks @winnielinnie . I appreciate the help and follow-ups. Do bugs get reported on github? I may need some help in describing this. :grinning:

1 Like
18

Bugs are posted on iXsystems Jira.

You can describe the bug as the middleware gets confused by pools/vdevs using non-partitioned, full drives. You can link them to this thread.

19

yeah it is a lot. i had to learn a lot of this myself to switch over. but now i know what to do more or less, it’s easier next time so it’s worth it :saluting_face:

20

Thanks @winnielinnie
I created a bug report on Jira. I’ve never used it before so it was drinking from a firehose again. :rofl:
Hopefully it worked OK.