veganbtw

When trying to enable tailscale in this container: Eventually: I think there's an issue with the container and the pre-defined tailscale script.

Well, the problem there is not knowing that plugin even exists, the terrible search, and outdated information / documentation. I didn't know about that plugin until you pointed it out. Keyword searching for "unraid subfolder sharing" brings up this reddit thread which links to an old unraid forum topic: https://www.reddit.com/r/unRAID/comments/ztwhmd/creating_smb_shares_for_subfolders/ Hard to like something you don't even know exists. At the very least they could get someone to maintain the official documentation: https://docs.unraid.net/ Oh, did you want to share a subfolder? Nope, can't be done. Here's a link to a plugin. Having to navigate through a bunch of old videos and forum posts to find new / recent information is just straight up bs. If this was free, sure, that's understandable. I mean, they could open up the docs to be edited by users and that might make it better, almost like a wiki. Anyway, just ranting a bit here that'll never get read by Limetech. Thanks for indulging me. I've already started migrating back to Windows Server because it just works™ and I don't need plugins to do basic things like sharing a subfolder. I'm too old to be paying companies for the privilege to beta test for them.

What does that even mean? When you make a USB, I just use the default recommended option, which is 7.2.3. Probably 7.2.4 now that it just released today.

It's basic stuff like this that should be built-in already which questions the design decisions of Limetech and whether or not there's a bunch of idiots behind the scenes. Stuff like Tmux should already be there. Unassigned devices plugin; why does this need to be installed separately? So many fundamentals missing that need to be manually added later is just dumb design.

It was the move operation using the built-in file manager that I couldn't cancel from above where it just shows "job running" but I can't view it anywhere. F me for using what's built into Unraid right?

This is insanely stupid design. There's no freaking way to kill the process. It just gets stuck like this. There's no task manager. What the hell kind of design is this server. This is so stupid.

7.2.3 current latest There's nothing down there. There really should be a dedicated button at the top right like the terminal icon.

This seems like a major oversight. I set a move job but I cannot see the progress anymore after minimizing the initial window. And I can't do anything else in the meantime other than cancel the job apparently if I click that option. Nothing shows up at the bottom where others have had a progress message.

Coming from Windows Server, if I move (cut+paste) files within the same disk, it's instant. If it's between different disks, it'll read and write all the files. I can't seem to figure out how to do this on unraid. The built-in file manager through the browser (when you explore the share) doesn't seem to work. Built-in midnight commander also doesn't seem to work. They always read and write when moving between shares on the same disk. Even when I move between the disk; ex. /mnt/disk1/share-a => /mnt/disk1/share-b. It doesn't make sense to have to do that on the same disk. Think I might've found my answer here: https://forums.unraid.net/topic/176113-moving-files-on-same-disk/ I'm using ZFS pools, apparently shares are individual datasets and they always have to copy. Need to figure out a better way to manage files, maybe put them all in 1 dataset and organize from there so it's instant. Update: This doesn't work as shares can't be assigned to subfolders / subdatasets. Well that's annoying.

Is there a way to delete the existing generated keys and re-generate new ones? 2 of my pools doesn't unlock with the auto-unlock; they show up as incorrect fs and need to be formatted. I need to delete the key, shut down the array, and then enter the passphrase manually for those 2 pools to work properly.

At the top right, there's a button to open a terminal window. I ran an rclone command and closed the window, thinking it'd continue and I could check back later. It actually cancelled the command and when I re-opened later, it's a new terminal window. Is there a way to have a persistent CLI terminal? Would using the rclone GUI docker container be a suitable alternative to not requiring a window open? Thanks

Found this post: https://forums.unraid.net/topic/184545-unable-to-increase-zfs-arc-cache/ Apparently you have to make this file: /boot/config/modprobe.d/zfs.conf with this content (for 32GB): options zfs zfs_arc_max=32000000000 (^ 9 zeroes) Was there a way to configure this through the GUI or is this the only way? Unraid's default has it set to 1/8 of available memory, or 12.5%. Is there a way to set it as a % rather than a fixed amount?

Yeah, after doing some more reading, seems to be using the same parity-to-space ratio as the original configuration and a re-writing of all data is required to fix it, might as well re-create the whole pool. 🤦‍♂️ Why even release a feature that doesn't even work properly. Would've preferred to not even be aware of this 'feature' altogether.

It's also on the main screen:

I have 3x10TB drives. I started as raidz1 with 2 disks, 10TB useable, which is correct. Once I expanded with another 10TB, only got ~5TB added of useable space. Raidz1 with 3x10TB should have 20TB useable space. It's scrubbing right now, I think it does that automatically after an expansion.

Regarding this last part, I had to change replace a motherboard with the same model and added a new GPU. Otherwise the rest of the hardware is the same; I transferred the storage controller and NIC over. When I tried to generate news keys, it just gives a big yellow banner saying: ERROR =========================================================== Prior to the hardware change, the auto-unlock worked fine. Update: I couldn't access the community apps section, apparently it was some issue with the time sync. When I checked the time setting, it was unsynchronized. Could not get it to sync with the time. Could've been a network issue, but it didn't make sense as I was connecting to it via the network. Ended up re-making a new USB and starting over. Seems to work now.

According to the docs: https://docs.unraid.net/unraid-os/using-unraid-to/manage-storage/file-systems/ XFS is "robust against corruption". What I've read is the unraid parity of XFS drives doesn't have a "self-healing"-like feature that a ZFS pool would have or bitrot protection. More realistically though, unclean shutdowns would be the most common issues I'd encounter. How would an XFS parity array fair against a ZFS mirrored pool? Actually, faulty cables and storage controllers would be another point of issue as well.

Hey, thanks for this plugin. There really needs to be more information like this in the official unraid docs: https://docs.unraid.net/unraid-os/system-administration/secure-your-server/securing-your-data/ Sifting through forum posts mixed with older/outdated methods is annoying.

I see in an uncoming version of Unraid that you can move the boot drive to internal drives that can be in a pool of multiple drives to offer redundancy, which is great news. Would this also be an option for USB flash drives? A pool of them so if one fails, it can still boot from another USB that's plugged in? I actually don't mind that it's USB based as long as there's some sort of redundancy. USB ports are never really used in a headless server anyway and internal storage slots are a premium resource.

I have 6 drives. Divided into 3 pools or 2 drives each. I can't seem to erase or remove pools without stopping everything. Is this normal or am I missing something? Coming from Storage Spaces (Windows Server) where the pools are independent of each other, this seems weird. Example: I have a pool for files and a pool for VMs. If I take down the VMs to add a disk or something, at least the files are still available to users during the downtime of the VMs.

Does this mean Unraid intends to make ZFS the default in the near future? Or is the current ZFS support still too immature? My goal is to avoid downtime and changing the filesystem in the future.

Coming from Windows Server ReFS and storage spaces where we use pools for everything. Trying to test and get an idea of the filesystems before the actual migration. I see some people complaining about BTRFS corruption seem to have bad hardware (RAM). I have ECC but those can also fail too, although much rarer. Is ZFS that much better for data integrity at the cost of performance? The pool will be for VMs and containers. BTRFS is still the default for pools, so there must be good reasoning behind it.

If anyone has transitioned from Windows Server to Unraid, I would be very interested in your experience and wisdom. Encryption. Is there an equivalent to Bitlocker? I haven't come across anything like it; being able to fully encrypt and decrypt without having to wipe the drive or start fresh. Recovery is simple with the recovery key provided and combining TPM with a password for security. The OS drive can also be encrypted to require a password and/or TPM on boot. The Unraid docs go into the XFS, ZFS, and BTRFS file systems a bit, but I'd like to know some real world implications for each so I don't choose the wrong one and then have to undo everything because I chose the wrong filesystem. It's simple with Windows, ReFS and NTFS. NTFS as the filesystem inside the virtual disks and those virtual disks are stored on ReFS formatted storage pool. ReFS has a file system layer of thin provisioning for virtual disks, so instead of writing to allocate space when creating a virtual disk, it's done instantly by reserving the space instead and data will be written later when there is actual data. This happens even on fixed size disks, so the size doesn't balloon like on thin-provisioned disks when a lot of data is written over. Is there something like that with either XFS, ZFS, or BTRFS? The scenario I'm looking to create it quite straightforward but may change as I don't think I'll need many virtual machines with the availability of containers. 2 storage pools, 1 for SSDs and 1 for HDDs. Virtual machines and virtual disks on each respective pool with encryption on the pool or the disks. 🙏