Unassigned Devices - Managing Disk Drives and Remote Shares Outside of The Unraid Array

chuckthetekkie · March 3, 2023

7 minutes ago, dlandon said:

That recommendation was from many years ago because is seemed to be an easy answer to just use the /mnt/disks/ or /mnt/remotes/ because UD applied a protection to prevent the rootfs from filling up if something was misconfigured. While it appears to work in most situations, it is a problem for you because your zfs disk is mounted at /mnt/disks/ before UD has installed. We are trying to re-think this to see if there might be a better answer.

If you postpone the zfs mount until after UD starts, you shouldn't get the reboot message.

How do I do that?

dlandon · March 3, 2023

2 minutes ago, chuckthetekkie said:

How do I do that?

I have no idea. I don't have any experience with the zfs plugin.

qw3r7yju4n · March 3, 2023

Would you have a recommended way of remounting it to retrieve the data. Or should I just reboot?

dlandon · March 3, 2023

1 minute ago, qw3r7yju4n said:

Would you have a recommended way of remounting it to retrieve the data. Or should I just reboot?

I would shut it down, check that cables are secure, boot it up, mount it and remove the data. You will probably have to reformat the disk. Why are you using ext4? XFS or BTRFS would be better choices for Unraid. NTFS if you want Windows read/write capabilities.

qw3r7yju4n · March 3, 2023

24 minutes ago, dlandon said:

I would shut it down, check that cables are secure, boot it up, mount it and remove the data. You will probably have to reformat the disk. Why are you using ext4? XFS or BTRFS would be better choices for Unraid. NTFS if you want Windows read/write capabilities.

Ty sir. My weekend project has just been selected

sonofdbn · March 3, 2023

On 3/2/2023 at 5:06 AM, dlandon said:

There may be something in the UD config file that is causing this issue. Do this:

Remove UD.

PM me the '/flash/config/plugins/unassigned.devices/unassigned.devices.cfg'.

Remove the '/flash/config/plugins/unassigned.devices/unassigned.devices.cfg' file.

Re-install UD.

Re-install your SSD device.

I carried out these steps and for a few hours everything was OK. But then after a short time the problem cropped up again: the UD part showed the animated logo and the GUI slowly died. Some docker containers couldn't be reached, others OK, VM OK.

Anything else I should do?

dlandon · March 3, 2023

1 hour ago, sonofdbn said:

I carried out these steps and for a few hours everything was OK. But then after a short time the problem cropped up again: the UD part showed the animated logo and the GUI slowly died. Some docker containers couldn't be reached, others OK, VM OK.

Anything else I should do?

Post diagnostics before you reboot.

dlandon · March 3, 2023

Also, when the UD GUI hangs, show the output of this command:

ls -la /tmp/unassigned.devices/

dlandon · March 3, 2023

There is a new feature in UD to help with plugins that need a mount point folder. UD now creates a mount folder '/mnt/addons'. This folder should be used to mount your plugin devices to prevent FCP warnings and keep your mounts separated from UD. It is recommended you switch to this new folder for your addon and custom mounts. You are welcome to leave the mounts at '/mnt/disks' or '/mnt/remotes'. Things will still work, but if you request support for UD issues, you will be asked to remove them first. One thing to keep in mind is there is a protection added to the '/mnt/addons' so a misconfigured app cannot fill the rootfs and crash the system. If you mount something at '/mnt/addons' before UD is installed, you will see a reboot message over the banner. This is because UD cannot apply the protection while a device is mounted at '/mnt/addons'. You will have to delay any mounts at '/mnt/addons' until UD has installed.

dlandon · March 3, 2023

Sometime ago, a file locking mechanism was added to UD to insure exclusive access when writing files. The idea is to prevent two processes from trying to write a configuration file at the same time. This insures that one process does not overwrite another processe's changes. File locking also applies to saving udev information on devices. The file lock was applied even when just reading the udev file. This created a lot of extra ramfs file system work that was not necessary because the refresh of UD does a lot of udev lookups to be sure the GUI is up to date. Because Linux is so poor at file locking, I had to use my own mechanism.

Where this is going is I suspect all that file locking may have been the source of 'stuck' UD where the Unraid wave was constantly showing. If this doesn't clear up the hangs, I have some debugging ideas so we can get to the source of the problem.

sonofdbn · March 4, 2023

8 hours ago, dlandon said:

Post diagnostics before you reboot.

Unfortunately I've already rebooted. Will grab diagnostics next time, but I'll be away for most of next week, and not having Plex running during that time will not go down well. I'll see if I can do a reboot before I leave, but no guarantees.

Edit: Oooh, a new shiny update of the plug-in. Have just installed it.

Edited March 4, 2023 by sonofdbn
Plug-in updated

csendre · March 4, 2023

On 3/2/2023 at 9:09 AM, Squid said:

You always want to map to a subdirectory in /tmp (eg: /tmp/plex) mapping directly to /tmp means that Plex has the ability to delete anything else within /tmp which may cause a ton of problems.

This seems to be the solution. Survived several recordings/transcodes now and all is good. Thank you!

chuckthetekkie · March 4, 2023

Another weird issue I am having is that some of the ZFS formatted devices do not have the PASSTHROUGH option but other ZFS ones do.

karamediagroup-diagnostics-20230303-2146.zip

dlandon · March 4, 2023

12 minutes ago, chuckthetekkie said:

Another weird issue I am having is that some of the ZFS formatted devices do not have the PASSTHROUGH option but other ZFS ones do.

karamediagroup-diagnostics-20230303-2146.zip 377.27 kB · 0 downloads

Turn on the pass through switch when the disk is not mounted. It should stick.

sonofdbn · March 4, 2023

13 hours ago, dlandon said:

Post diagnostics before you reboot.

The UD GUI hung again. I'm on UD version 2023.03.03. The diagnostics are attached.

Unfortunately I didn't read the next post at first. So I tried to stop the array using

emcmd cmdStop=Stop

Waited a while and looks like some services stopped (like VMs) but I couldn't get into the GUI - perhaps I should have expected this. Next I ran "powerdown" and heard some beeps but the array didn't shut down.

Then I saw that next post, so I ran

ls -la /tmp/unassigned.devices/

and this is the result:

total 4
drwxr-x---  5 root root 120 Mar  4 14:33 ./
drwxrwxrwt 17 root root 380 Mar  4 14:36 ../
-rwxrwx---  1 root root 634 Mar  4 09:48 add-smb*
drwxr-xr-x  2 root root 100 Mar  4 09:48 config/
drwxr-xr-x  2 root root  40 Mar  4 00:41 logs/
drwxr-xr-x  2 root root  40 Mar  4 00:41 scripts/

tower-diagnostics-20230304-1425.zip

dlandon · March 4, 2023

5 hours ago, sonofdbn said:
The UD GUI hung again. I'm on UD version 2023.03.03. The diagnostics are attached.

Unfortunately I didn't read the next post at first. So I tried to stop the array using
emcmd cmdStop=Stop
Waited a while and looks like some services stopped (like VMs) but I couldn't get into the GUI - perhaps I should have expected this. Next I ran "powerdown" and heard some beeps but the array didn't shut down.
5 hours ago, sonofdbn said:
Then I saw that next post, so I ran
ls -la /tmp/unassigned.devices/
and this is the result:
total 4
drwxr-x---  5 root root 120 Mar  4 14:33 ./
drwxrwxrwt 17 root root 380 Mar  4 14:36 ../
-rwxrwx---  1 root root 634 Mar  4 09:48 add-smb*
drwxr-xr-x  2 root root 100 Mar  4 09:48 config/
drwxr-xr-x  2 root root  40 Mar  4 00:41 logs/
drwxr-xr-x  2 root root  40 Mar  4 00:41 scripts/
tower-diagnostics-20230304-1425.zip 171.77 kB · 0 downloads

You have some page faults?

Mar  4 13:00:48 Tower kernel: BUG: kernel NULL pointer dereference, address: 00000000000000b6
Mar  4 13:00:48 Tower kernel: #PF: supervisor read access in kernel mode
Mar  4 13:00:48 Tower kernel: #PF: error_code(0x0000) - not-present page
Mar  4 13:00:48 Tower kernel: PGD 0 P4D 0 
Mar  4 13:00:48 Tower kernel: Oops: 0000 [#1] PREEMPT SMP PTI
Mar  4 13:00:48 Tower kernel: CPU: 8 PID: 1532 Comm: qbittorrent-nox Not tainted 5.19.17-Unraid #2
Mar  4 13:00:48 Tower kernel: Hardware name: Supermicro Super Server/X10SDV-TLN4F, BIOS 2.0c 06/27/2019
Mar  4 13:00:48 Tower kernel: RIP: 0010:folio_try_get_rcu+0x0/0x21
Mar  4 13:00:48 Tower kernel: Code: e8 9d fd 67 00 48 8b 84 24 80 00 00 00 65 48 2b 04 25 28 00 00 00 74 05 e8 c1 35 69 00 48 81 c4 88 00 00 00 5b c3 cc cc cc cc <8b> 57 34 85 d2 74 10 8d 4a 01 89 d0 f0 0f b1 4f 34 74 04 89 c2 eb
Mar  4 13:00:48 Tower kernel: RSP: 0000:ffffc90008d2bcc0 EFLAGS: 00010246
Mar  4 13:00:48 Tower kernel: RAX: 0000000000000082 RBX: 0000000000000082 RCX: 0000000000000082
Mar  4 13:00:48 Tower kernel: RDX: 0000000000000001 RSI: ffff8882764e4000 RDI: 0000000000000082
Mar  4 13:00:48 Tower kernel: RBP: 0000000000000000 R08: 0000000000000028 R09: ffffc90008d2bcd0
Mar  4 13:00:48 Tower kernel: R10: ffffc90008d2bcd0 R11: ffffc90008d2bd48 R12: 0000000000000000
Mar  4 13:00:48 Tower kernel: R13: ffff888046fcf2f8 R14: 0000000000237eeb R15: ffff888046fcf300
Mar  4 13:00:48 Tower kernel: FS:  00001542112ad6c0(0000) GS:ffff88885fc00000(0000) knlGS:0000000000000000
Mar  4 13:00:48 Tower kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Mar  4 13:00:48 Tower kernel: CR2: 00000000000000b6 CR3: 00000003477a8003 CR4: 00000000003726e0
Mar  4 13:00:48 Tower kernel: DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
Mar  4 13:00:48 Tower kernel: DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Mar  4 13:00:48 Tower kernel: Call Trace:
Mar  4 13:00:48 Tower kernel: <TASK>
Mar  4 13:00:48 Tower kernel: __filemap_get_folio+0x98/0x1ff
Mar  4 13:00:48 Tower kernel: filemap_fault+0x6e/0x524
Mar  4 13:00:48 Tower kernel: __do_fault+0x30/0x6e
Mar  4 13:00:48 Tower kernel: __handle_mm_fault+0x9a5/0xc7d
Mar  4 13:00:48 Tower kernel: handle_mm_fault+0x113/0x1d7
Mar  4 13:00:48 Tower kernel: do_user_addr_fault+0x36a/0x514
Mar  4 13:00:48 Tower kernel: exc_page_fault+0xfc/0x11e
Mar  4 13:00:48 Tower kernel: asm_exc_page_fault+0x22/0x30
Mar  4 13:00:48 Tower kernel: RIP: 0033:0x154216b7504d
Mar  4 13:00:48 Tower kernel: Code: 00 00 00 00 00 66 66 2e 0f 1f 84 00 00 00 00 00 66 66 2e 0f 1f 84 00 00 00 00 00 66 90 f3 0f 1e fa 48 89 f8 48 83 fa 20 72 33 <c5> fe 6f 06 48 83 fa 40 0f 87 b5 00 00 00 c5 fe 6f 4c 16 e0 c5 fe
Mar  4 13:00:48 Tower kernel: RSP: 002b:00001542112ab2c8 EFLAGS: 00010202
Mar  4 13:00:48 Tower kernel: RAX: 0000153640058f60 RBX: 00001542112ac308 RCX: 00001542112ab500
Mar  4 13:00:48 Tower kernel: RDX: 0000000000004000 RSI: 000015088f8ebfe1 RDI: 0000153640058f60
Mar  4 13:00:48 Tower kernel: RBP: 0000000000000000 R08: 0000000000000001 R09: 0000000000000000
Mar  4 13:00:48 Tower kernel: R10: 0000000000000008 R11: 0000000000000246 R12: 0000000000000000
Mar  4 13:00:48 Tower kernel: R13: 0000153640004b80 R14: 0000000000000001 R15: 0000557bf3268f70
Mar  4 13:00:48 Tower kernel: </TASK>
Mar  4 13:00:48 Tower kernel: Modules linked in: xt_mark tcp_diag udp_diag inet_diag xt_CHECKSUM ipt_REJECT nf_reject_ipv4 ip6table_mangle ip6table_nat iptable_mangle vhost_net vhost vhost_iotlb tap tun veth macvlan xt_nat xt_tcpudp xt_conntrack xt_MASQUERADE nf_conntrack_netlink nfnetlink xfrm_user xt_addrtype iptable_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 br_netfilter xfs md_mod ip6table_filter ip6_tables iptable_filter ip_tables x_tables bridge stp llc ixgbe xfrm_algo mdio igb x86_pkg_temp_thermal intel_powerclamp mxm_wmi coretemp ast drm_vram_helper drm_ttm_helper ttm kvm_intel drm_kms_helper kvm crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel aesni_intel ipmi_ssif crypto_simd cryptd rapl intel_cstate drm intel_uncore mpt3sas backlight i2c_i801 agpgart i2c_smbus ahci intel_pch_thermal i2c_algo_bit input_leds joydev acpi_ipmi libahci led_class raid_class syscopyarea sysfillrect i2c_core sysimgblt scsi_transport_sas fb_sys_fops ipmi_si wmi acpi_pad button unix
Mar  4 13:00:48 Tower kernel: [last unloaded: xfrm_algo]
Mar  4 13:00:48 Tower kernel: CR2: 00000000000000b6
Mar  4 13:00:48 Tower kernel: ---[ end trace 0000000000000000 ]---
Mar  4 13:00:48 Tower kernel: RIP: 0010:folio_try_get_rcu+0x0/0x21
Mar  4 13:00:48 Tower kernel: Code: e8 9d fd 67 00 48 8b 84 24 80 00 00 00 65 48 2b 04 25 28 00 00 00 74 05 e8 c1 35 69 00 48 81 c4 88 00 00 00 5b c3 cc cc cc cc <8b> 57 34 85 d2 74 10 8d 4a 01 89 d0 f0 0f b1 4f 34 74 04 89 c2 eb
Mar  4 13:00:48 Tower kernel: RSP: 0000:ffffc90008d2bcc0 EFLAGS: 00010246
Mar  4 13:00:48 Tower kernel: RAX: 0000000000000082 RBX: 0000000000000082 RCX: 0000000000000082
Mar  4 13:00:48 Tower kernel: RDX: 0000000000000001 RSI: ffff8882764e4000 RDI: 0000000000000082
Mar  4 13:00:48 Tower kernel: RBP: 0000000000000000 R08: 0000000000000028 R09: ffffc90008d2bcd0
Mar  4 13:00:48 Tower kernel: R10: ffffc90008d2bcd0 R11: ffffc90008d2bd48 R12: 0000000000000000
Mar  4 13:00:48 Tower kernel: R13: ffff888046fcf2f8 R14: 0000000000237eeb R15: ffff888046fcf300
Mar  4 13:00:48 Tower kernel: FS:  00001542112ad6c0(0000) GS:ffff88885fc00000(0000) knlGS:0000000000000000
Mar  4 13:00:48 Tower kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Mar  4 13:00:48 Tower kernel: CR2: 00000000000000b6 CR3: 00000003477a8003 CR4: 00000000003726e0
Mar  4 13:00:48 Tower kernel: DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
Mar  4 13:00:48 Tower kernel: DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400

I don't know enough about this to offer much help, but I sure would do a memory test.

5 hours ago, sonofdbn said:
Then I saw that next post, so I ran
ls -la /tmp/unassigned.devices/
and this is the result:

What I was looking for was lock files that were orphaned that might cause UD to hang, but there are none.

sonofdbn · March 4, 2023

2 hours ago, dlandon said:

You have some page faults?

I don't know enough about this to offer much help, but I sure would do a memory test.

Thanks for this. No time to do the memtest now, but will get on to it when I'm back from my travels.

akazu · March 4, 2023

Hello Community

I have mounted a hard disk under Unassigned Devices and enabled share for the disk and partition.

However, when I go to the share from my Windows PC "\Unraid"\ I do not see the share.

In Unraid I see the data on the disk.

Is there still a hidden function somewhere which I have to deactivate? How can I access the share?

Thanks for your help

burgsth · March 4, 2023

22 minutes ago, akazu said:

Hello Community

I have mounted a hard disk under Unassigned Devices and enabled share for the disk and partition.

However, when I go to the share from my Windows PC "\Unraid"\ I do not see the share.

In Unraid I see the data on the disk.

Is there still a hidden function somewhere which I have to deactivate? How can I access the share?

Thanks for your help

I have the same problem since todays plugin update. Share feature seems to be broken.

dlandon · March 4, 2023

27 minutes ago, akazu said:

Thanks for your help

4 minutes ago, burgsth said:

I have the same problem since todays plugin update. Share feature seems to be broken.

Go to UD Settings and set the SMB Security for the way you want to share the UD devices.

akazu · March 4, 2023

3 minutes ago, burgsth said:

I have the same problem since todays plugin update. Share feature seems to be broken.

for testing i have included it in Krusader as follows. However, this does not work under Windows.

dlandon · March 4, 2023

3 minutes ago, akazu said:

for testing i have included it in Krusader as follows. However, this does not work under Windows.

Did you set the UD Settings->SMB Security to something other than 'No'? And give up to two minutes for the share to be active?

akazu · March 4, 2023

11 minutes ago, dlandon said:

Thanks for your help

Thanks for your work and help

13 minutes ago, dlandon said:

Go to UD Settings and set the SMB Security for the way you want to share the UD devices.

Oha, I didn't know that. That has solved my problem, thank you very much :-)

akazu · March 4, 2023

19 minutes ago, dlandon said:

Did you set the UD Settings->SMB Security to something other than 'No'? And give up to two minutes for the share to be active?

I am new to Unraid and I was not aware that I need to set SMB sharing in the UD settings. Thank you, I have learned again and it works now

Edited March 4, 2023 by akazu

dlandon · March 4, 2023

UD used to share devices 'Public' when SMB Security was set to 'No' as a convenience. This was changed in June of 2022 for security reasons. Per release notes:

2022.06.19
Add: 'Public' setting to SMB Security settings to share devices public. The 'No' setting used to set 'Public' sharing.

SMB Security now has to be specifically set to how you want to share all UD devices - Public, Private, or Private Hidden. The 'Share' switch on each device lets you control if that particular device is shared.

For you newbees, always assume in Unraid that if you want something shared with SMB or NFS, you have to specifically set up how it will be shared so you always have control over your share security. Just setting a 'Share' switch does not specify how a device will be shared.

For those of you old timers just now updating and seeing this issue, how about you update more often than 7 or 8 months, and read the latest release notes so I don't have to explain this issue over and over like I have in the last few days. Set your system to either automatically update plugins, or at least notify you of a plugin update and keep plugins updated!

Unassigned Devices - Managing Disk Drives and Remote Shares Outside of The Unraid Array

Recommended Posts

Link to comment

Top Posters In This Topic

Popular Days

Top Posters In This Topic

Popular Days

Popular Posts

dlandon

dlandon

dlandon

Posted Images

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Join the conversation