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

[shingaling]

thanks @dlandon 

I set new permissions (on the synology serving the NFS shares)

I tried to lower the protocol (from 4.1 to 3)

 

The weirdest thing:

A reboot on my second unraid server doesn't auto-mount any NFS shares (see attached diagnostics)

A reboot on my first unraid server auto-mounts only one of the 2 servers that have nfs shares (issue as described above by you)

 

So I am stuck. I mean: Everything will be easy, as long there is no power outage. If there is, all wheels stand still, and I have to manually mount each share.

 

Edited March 9 by shingaling

[dlandon]

1 hour ago, shingaling said:

The weirdest thing:

A reboot on my second unraid server doesn't auto-mount any NFS shares (see attached diagnostics)

A reboot on my first unraid server auto-mounts only one of the 2 servers that have nfs shares (issue as described above by you)

Your wait timer before mounting remote shares is 30 seconds.  Make that several minutes and see if that's enough time for things to get settled and allow the mounts.

 

[shingaling]

Hey @dlandon 

I found (a weird noob) solution:

• On each Synology serving the NFS-Shares I created a new share (wakeywakey)
• Then in UD i added this share (wakeywakey) as a remote NFS share, but not using the name of the (Synology) Server but instead using the IP-Adress (192.168.xx.xxx:/volume1/wakeywakey)

Now when I reboot the system (or it reboots after a power failure), it mounts all shares with no issue and no delay (i set the delay down to 12secs, guess I could go to the standard of 5sec.

 

This works for me ... hope this helps anyone having the same issue.

Edited March 9 by shingaling
explain better

[dlandon]

5 hours ago, shingaling said:

I found (a weird noob) solution:

That's not a weird solution.  What that means is that the name resolution is taking a while and using the IP address that doesn't rely on name resolution gets past that.

[shingaling]

1 hour ago, dlandon said:

What that means is that the name resolution is taking a while and using the IP address that doesn't rely on name resolution gets past that.

Actually name resolution doesn’t happen. I had it idle for 20+ minutes without Unraid resolving the name of the NFS-Sharing-Servers. Only after this „trick“ with calling their ip‘s first, did the nameresolution happen at all. 
 

at least now i can let it run and not worry that once the power fails, all connectivity is gone after reboot. 
 

thanks for your help @dlandon

👍🏻

Edited March 9 by shingaling

[d3m3zs]

Hi! I already posted my question to ZFS plugin, but author told me it is issue of Unassigned device plugin.

How can I fix it?

Unassigned ZFS disks automatically spin up when I open Main tab on Unraid, all other ZFS disks (in array) continue sleeping.

 

Edited March 12 by d3m3zs

[dlandon]

16 minutes ago, d3m3zs said:

How can I fix it?

"ZFS file systems have built-in mechanisms such as periodic scrubs, snapshots, and metadata updates that can prevent disks from entering standby mode. These operations are designed to maintain data integrity and can cause disk activity even when the system is idle, keeping the disks from entering low-power states."

 

That being said, UD has to check unassigned disks when it polls for various status and may be triggering a zfs disk operation.  It acts differently from a pool disk because UD disks come and go and pool disks are static and only change when the array is started or stopped.  If the UD disk is not mounted, you can detach it and that will put it into a standby mode.  If it is always mounted, I would suggest moving the disk to a pool disk.

 

I notice some disk read operations when a zfs disk is unmounted, and I may put a little time into understanding what UD is doing that might cause that.

[d3m3zs]

26 minutes ago, dlandon said:

"ZFS file systems have built-in mechanisms such as periodic scrubs, snapshots, and metadata updates that can prevent disks from entering standby mode. These operations are designed to maintain data integrity and can cause disk activity even when the system is idle, keeping the disks from entering low-power states."

 

That being said, UD has to check unassigned disks when it polls for various status and may be triggering a zfs disk operation.  It acts differently from a pool disk because UD disks come and go and pool disks are static and only change when the array is started or stopped.  If the UD disk is not mounted, you can detach it and that will put it into a standby mode.  If it is always mounted, I would suggest moving the disk to a pool disk.

 

I notice some disk read operations when a zfs disk is unmounted, and I may put a little time into understanding what UD is doing that might cause that.

Thanks, by the way, BTRFS has the same features, but never spin up (I have also BTRFS on unassigned disks)

[dlandon]

Just now, d3m3zs said:

Thanks, by the way, BTRFS has the same features, but never spin up (I have also BTRFS on unassigned disks)

I was looking at the code and may have found an issue.  Standby.

[dlandon]

30 minutes ago, d3m3zs said:

Thanks, by the way, BTRFS has the same features, but never spin up (I have also BTRFS on unassigned disks)

I think I have a fix.  UD was doing a zpool operation when it was not necessary and that will spin up a disk.  I'll do some testing and if it works out, the fix will be in the next release of UD.

[runawaydevil]

Hello, im noob here

I just installed UD plus and when I set destructive mode and apply it carries me for update.php and dont activate, is something im doing wrong?

[dlandon]

45 minutes ago, runawaydevil said:

Hello, im noob here

I just installed UD plus and when I set destructive mode and apply it carries me for update.php and dont activate, is something im doing wrong?

Can you post the screen shot of the php warning, or go to Tools->PHP Settings, and show the warnings,

[dlandon]

On 3/12/2024 at 5:26 PM, d3m3zs said:

Thanks, by the way, BTRFS has the same features, but never spin up (I have also BTRFS on unassigned disks)

Update UD and see if it is fixed.

[d3m3zs]

1 hour ago, dlandon said:

Update UD and see if it is fixed.

Thank you, updated and it seems working, unassigned ZFS disks do not spin up anymore.  🤝

[brambo23]

So I'm currently running 6.9.2 and tried to install your latest version and it requires 6.11.  is there any chance you have an old version I can use to get my array back up and running?  I'm not seeing it in community apps or your github

[dlandon]

8 hours ago, brambo23 said:

So I'm currently running 6.9.2 and tried to install your latest version and it requires 6.11.  is there any chance you have an old version I can use to get my array back up and running?  I'm not seeing it in community apps or your github

Your best bet is to upgrade to 6.11.  The other option is to copy the /boot/config/plugins/ folder from a backup and reboot.

[Sionree]

Hello everyone.

I've successfully used remote shares for several months.

After a blackout, I haven't been able to get it working again.

"Search for servers" works correctly, identifying the PC with Win10 containing the folder; however, "load shares" doesn't find any shares.

I've uninstalled and reinstalled without success.

I've installed a "free trial" version on another PC, and remote shares works correctly there.

Thanks to anyone who can advise me.

unraidserver-diagnostics-20240317-1427.zip

[dlandon]

15 minutes ago, Sionree said:

Hello everyone.

I've successfully used remote shares for several months.

After a blackout, I haven't been able to get it working again.

"Search for servers" works correctly, identifying the PC with Win10 containing the folder; however, "load shares" doesn't find any shares.

I've uninstalled and reinstalled without success.

I've installed a "free trial" version on another PC, and remote shares works correctly there.

Thanks to anyone who can advise me.

unraidserver-diagnostics-20240317-1427.zip 200.54 kB · 0 downloads

You have to enter the correct credentials in order to list the shares.  As a test, when you get to the part where yoiu enter a share, enter a share manually.  Then mount the share and let's see if an error shows.

[Sionree]

 

Left: Test server with the automatically found and mounted folder.

Right: My current server, with the folder manually entered, cannot be mounted.

Edited March 17 by Sionree

[dlandon]

14 minutes ago, Sionree said:

 

Left: Test server with the automatically found and mounted folder.

Right: My current server, with the folder manually entered, cannot be mounted.

Post diagnostics from the server that is failing.

[Sionree]

unraidserver-diagnostics-20240317-1622.zip

[dlandon]

A few things to try:

• You are using tailscale.  Check your setup and be sure the 'SARAH' server is seen.  Try pinging it from the server that is not showing it as mountable.  Ping both the name and IP address.
• Try using the IP address of 'SARAH' in UD instead of the name.  The issue may be name lookup.

[Sionree]

Thank you for the time you dedicated to me, it's working perfectly now. The SARAH PC no longer had Tailscale installed, but it still appeared in the GUI, which was causing issues. Once removed from the Tailscale network, local shares started working again. Have a great Sunday.

[DS18]

I have 2 unraid devices on the same local network, 1 is more of a client serving up all my apps and the other is a storage device. I connect from the client to the storage device over NFS and it connects fine but after some time, i get an Invalid Path issue on the client. The mount says its mounted still so i have to unmount and remount. I have it set to automount but i still need to remount and i typically notice this when i cannot connect to some apps on my client that depend on the data from the storage device. I have security turned on on the storage device with this: `192.168.1.68/24(sec=sys,rw,insecure,anongid=100,anonuid=99,no_root_squash)` Security is set to private, export no. Nothing special on the client device outside of 2 apps (plex and immich) have the share mounted as a volume for content.

 

Immich Mount looks like this:

` - ${UPLOAD_LOCATION}/immich:/usr/src/app/upload`

UPLOAD_LOCATION=/mnt/remotes/TOWER_media/Photos

 

Plex is slightly different since its in a different compose file:

`- /mnt/remotes/TOWER_media:/data:rw,slave`

 

Unsure if i need the slave part. 

 

Does anyone have any idea why my share keeps dropping after some unknown time/reason?

[dlandon]

1 hour ago, DS18 said:

I have 2 unraid devices on the same local network, 1 is more of a client serving up all my apps and the other is a storage device. I connect from the client to the storage device over NFS and it connects fine but after some time, i get an Invalid Path issue on the client. The mount says its mounted still so i have to unmount and remount. I have it set to automount but i still need to remount and i typically notice this when i cannot connect to some apps on my client that depend on the data from the storage device. I have security turned on on the storage device with this: `192.168.1.68/24(sec=sys,rw,insecure,anongid=100,anonuid=99,no_root_squash)` Security is set to private, export no. Nothing special on the client device outside of 2 apps (plex and immich) have the share mounted as a volume for content.

 

Immich Mount looks like this:

` - ${UPLOAD_LOCATION}/immich:/usr/src/app/upload`

UPLOAD_LOCATION=/mnt/remotes/TOWER_media/Photos

 

Plex is slightly different since its in a different compose file:

`- /mnt/remotes/TOWER_media:/data:rw,slave`

 

Unsure if i need the slave part. 

 

Does anyone have any idea why my share keeps dropping after some unknown time/reason?

Sounds like you might be getting stale file handles.  Post diagnostics.