grana

argh, ufs is not from command line, needs a screen. Is there any way to use VNC or something to emulate a screen and connect to it from my laptop?

and the output is cannot import 'disk3': one or more devices is currently unavailable which I guess is not good news. I'll give a final try with UFS Explorer.

thanks, I will report back once it's done (1 day?)

zpool import disk3 cannot import 'disk3': insufficient replicas Destroy and re-create the pool from a backup source. and zdb -U /data/zfs/zpool.cache -eCC disk3 Configuration for import: vdev_children: 1 version: 5000 pool_guid: 2087549526651695302 name: 'disk3' state: 0 vdev_tree: type: 'root' id: 0 guid: 2087549526651695302 children[0]: type: 'disk' id: 0 guid: 18181126314476381217 whole_disk: 0 metaslab_array: 256 metaslab_shift: 34 ashift: 12 asize: 14000497885184 is_log: 0 create_txg: 4 expansion_time: 1712610252 path: '/dev/mapper/md3p1' devid: 'dm-uuid-CRYPT-LUKS2-291a6f7d67b049978cdfc4ef46119047-md3p1' phys_path: '/dev/disk/by-uuid/2087549526651695302' load-policy: load-request-txg: 18446744073709551615 load-rewind-policy: 2 MOS Configuration: version: 5000 name: 'disk3' state: 0 txg: 1278326 pool_guid: 2087549526651695302 errata: 0 hostname: 'pescaflix' com.delphix:has_per_vdev_zaps vdev_children: 1 vdev_tree: type: 'root' id: 0 guid: 2087549526651695302 create_txg: 4 children[0]: type: 'disk' id: 0 guid: 18181126314476381217 path: '/dev/mapper/md3p1' whole_disk: 0 metaslab_array: 256 metaslab_shift: 34 ashift: 12 asize: 14000497885184 is_log: 0 create_txg: 4 expansion_time: 1712610252 com.delphix:vdev_zap_leaf: 129 com.delphix:vdev_zap_top: 130 features_for_read: com.delphix:hole_birth com.delphix:embedded_data

with the array started in normal mode (non maintenance): zpool import pool: disk3 id: 2087549526651695302 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: disk3 ONLINE md3p1 ONLINE also attaching new diags. Thanks pescaflix-diagnostics-20240414-1354.zip

zpool import pool: cache id: 15828626886734704336 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: cache ONLINE sdb1 ONLINE pool: disk3 id: 2087549526651695302 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: disk3 ONLINE md3p1 ONLINE pool: disk2 id: 11012034467400980073 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: disk2 ONLINE md2p1 ONLINE pool: disk1 id: 4272399308488201852 state: ONLINE action: The pool can be imported using its name or numeric identifier. config: disk1 ONLINE md1p1 ONLINE this is after starting the array in maintenance mode. With the array stopped, only cache entry is shown.

thanks, appreciate your help. Sorry for not pointing out the device is/was zfs.

The manual says that documentation for ZFS should be added, but it is not

Hi, ~20 hours ago power disconnected from my unraid machine. When booting up again, as expected, the system started a parity check, which is still running. I was not particularly concerned with data validity as most of the times my disks are spun down, and only the cache is being written. Though, I just noticed something pretty bad: all the files created in the past 8 months or so, seem to be lost, unless for some miracle they will reappear when the parity check ends. This behavior is different from anything I've experienced in the past. Another thing I'm noticing is that one of the disk is showing "no file system" and is not being mounted. But, on the other hand, that disk seem to be being used for rebuilding parity, thus also wiping my chance of getting my data back through that. Attaching diagnostic files. Please help. pescaflix-diagnostics-20240413-2203.zip

Thank you! That worked.

yes

One of my user shares, pictured below, is configured to be on Array only. Somehow, though, there's an empty directory `/vr` on my cache drive, which I cannot remove in any way. Any help please? rm: cannot remove 'vr': Device or resource busy

I've recently moved from QNAP to Unraid, and there is one thing that I'm missing. I guess all users have some high priority media / documents that want to be as safe as possible, and some other content that has lower priority. Say for example I have a Windows VM; I probably care a lot about the content of the VM itself, or the configuration, but the original ISO downloaded from microsoft.com? Not so much. I don't really need those x GB to be duplicated or even triplicated. As a noob user (I purchased the license yesterday, after 15 days of evaluation) what I am doing is having 2 disks in my array (1 + 1 parity), and another disk as pool device. Then, for example, all my docker images are configured to have configurations stored on the array disk (`/mnt/user`), but the scrap files on `/mnt/my_pool_disk`. Does it make sense? Now, to QNAP. QNAP has this nice feature where you can 1. partition the disk. Partitions are not fixed size, but they can grow or shrink based on the content. 2. for each partition, decide what cloning / redundancy strategy you want to adopt, and where the data should be cloned to (ie, one disk to another). Why does it matter? Well, it's simple. Say I have 3 disks, 4 TB each. The files I want to keep with parity are 500 GB. With QNAP, my total available storage is going to be 4 TB + 4 TB + 4 TB - 500 GB = 11.5 TB (500 GB of "wasted" duplicated storage). With Unraid my total storage is 8 TB (one parity disk, one disk in the array, one outside the array (or both of them into the array)). Basically, it would no longer be a parity disk, but a parity partition, calculating parity bits for a subset of the partitions on the disks. Thanks and keep up the great work

I want to add my take on this, thanking all the contributors of this thread but in particular @bonienl. My passkey is a simple password, not an image, and I didn't feel comfortable in leaving it in plain text anywhere, even if just on my local network. My idea is to encrypt the keyfile using an encryption key that UNRAID has readily available but that I don't have to memorize. Also, I feel like mounting a remote path is overkill, I'd rather download the keyfile over scp. So, I put the keyfile on a USB key on my router; for some reason the router supports SSH, but not sftp; but supports authentication by public key. So here's the script #!/bin/bash -xv if [[ ! -e /root/keyfile ]]; then # this is the equivalent of scp remote:source dest, using the host ssh key. The router does not support scp ssh -i /boot/config/ssh/ssh_host_rsa_key my.router.local "cat <path_to_key>/unraid_encrypted_keyfile.7z" > keyfile.7z # 7z (part of nerdtools) decompresses the file using the arg `-p` as password. The password for the decompression is the output of `cut -d' ' -f2 /boot/config/ssh/ssh_host_rsa_key.pub`, ie the host (unraid) public key. 7z e -p`cut -d' ' -f2 /boot/config/ssh/ssh_host_rsa_key.pub` keyfile.7z rm keyfile.7z mv keyfile /root/keyfile fi I'm pretty happy with the solution. The only thing I don't like, but I guess it's minor, is that `7z` is really slow, and it takes maybe 1 second to decrypt the file (this is really surprising honestly). Given that I don't care about the compression but just the encryption, I'm wondering what better tools I should use: `gpg` is apparently not part of nerd tools any more?