arretx

That link is chock full of malicious garbage. What happened to the actual guide?

My 1TB Cache pool shows 583GB of space used on the Main page. If I list files via terminal, I only get 387G (du -h). What would cause this?

As long as I've been doing this tippity tap stuff in a terminal, I've never asked the following question: If I run a bash script from the terminal, do I need to leave the terminal window open? Will the script continue to run if I close the terminal session?

Back in business. I skipped a step. Forgot to take the array offline to re-assign the disks to the cache pool

The server restarted, the array started, parity check began, and I noticed I had no VM's and no Docker containers again. This time, I looked at my Main page to find two SSD devices under "Disk Devices" labeled Dev 1 and Dev 2 that were unmounted. At this point, I don't know what to do short of ensuring one of the mirrored drives is completely backed up. Any help would be appreciated. Diagnostics attached. tower-diagnostics-20231205-1115.zip

When I look at my shares, and I click on /appdata, it shows that the contents are found on both `cache` and `disk5`. The /appdata share is set to Prefer cache, which would make me think that all of the important appdata files are actually on the cache disk. When I look at the contents of cache and disk5, I see nearly identical files and folders. There may have been, at one point, where, due to the potential failure of my cache pool, I changed all of my Prefer shares to Yes to force the mover to move the files off of the cache temporarily. The setting "Prefer" states that "Mover transfers files from array to cache" and "Yes" is the opposite, "Mover transfers files from cache to array." Under what circumstance might I have inadvertently duplicated content in the array that no longer needs to be there because it has been "moved" back to the cache? Are we actually moving files, or are we copying files when we employ the Cache settings? Thanks!

Intuitively, I would expect that if I set a share's cache setting from something other than "No" to "No" that the files would be moved off of the cache, share folder name and all. What I have found is that I have empty folders in my Cache Pool that may have once been shares that were cached, but are now set to "No." An example on my system would be a share called "Media." There's a folder on my cache pool called Media which is empty. The share "Media" is allocated to any available disk, but not the cache, and thus, no files show that they're on the cache. Is it safe for me to remove the folder that remains on the Cache Pool for this share?

Update, I just noticed that the 1TB cache pool is full (which isn't typically the case) and the /system/docker and /system/libvirt folders are located there as /system is marked as "prefer" on the cache settings. I imagine the cache being full is causing this problem, but I'm not certain. The bigger problem is that I'm unable to figure that out until the Parity check is done? That would mean my services are dead in the water for another 9 hours. Perhaps I'm not managing this system well, but 50+ hours to move critical data to healthy disks and then another 12 hours for parity check is a long time to not have critical services.

I also have this problem after replacing a disk. Everything that should be in /etc/libvirt is in /etc/libvirt- but I noticed that this /etc/libvirt- folder was created on October 6, and the problem started today.

The first thing I did when I had problems with Disk 1 (md1) was to run xfs_repair on the disk, ultimately requiring the -L switch. It came back online and I took the opportunity to use unBalance to move everything from that disk to the other disks of the array. It took about 50 hours and is complete. After that I shut down the system, replaced the drive, fired up the system, allocated it, formatted it, and parity is rebuilding data. What I'm not clear on is why both the Docker Service failed to start and none of the Virtual Machines are there anymore. I looked at the libvirt.log and found this: `2023-11-09 22:48:00.248+0000: 6006: error : catchXMLError:1015 : /etc/libvirt/qemu/Home Assistant.xml:1: Document is empty (null)` So I checked that path and found that there's nothing in /etc/libvirt (I get "/bin/ls: reading directory '.': Input/output error") if I list files. There is however, a path to /etc/libvirt- which contains the qemu folder, etc. Should I rename the hyphen folder to /etc/libvirt after removing /etc/libvirt? What about the Docker Service? Very confused at this point. tower-diagnostics-20231109-1806.zip

I just ran `xfs_repair -v /dev/md1` and was presented with this: Phase 1 - find and verify superblock... - reporting progress in intervals of 15 minutes - block cache size set to 3019504 entries Phase 2 - using internal log - zero log... zero_log: head block 439743 tail block 439417 ERROR: The filesystem has valuable metadata changes in a log which needs to be replayed. Mount the filesystem to replay the log, and unmount it before re-running xfs_repair. If you are unable to mount the filesystem, then use the -L option to destroy the log and attempt a repair. Note that destroying the log may cause corruption -- please attempt a mount of the filesystem before doing this. Destroying the log sounds bad. Should I do this?

Situation: Today, I was adding some power circuits to my house, so before I cut the mains, I shut down Unraid. Hours later, as I sit down to work on a few things, I notice that when I login, there's no array running due to missing cache drive. I've been using two NVMe drives, both 1TB, in a mirrored configuration. I used the reboot feature and it came back with the same problem. So, I shut the box down completely and powered it back up (something I had already done earlier to accommodate the main power being taken offline.) This time, it booted with the 2nd cache disk available, but until I unassigned the first and put the 2nd in the first position, I wasn't able to start the array. So, I started the array and quickly noticed that Disk 1 of the array reports "Unmountable: Wrong or no file system." This system has 2 parity drives each at 8TB and 5 additional drives in the array for data, etc. 4 are 8TB and 1 is 500GB. One of those 8TB drives is Disk 1. Have I lost data at this point? If so, how do I know what was on Disk 1? If I replace the disk, will the Array rebuild the data that was on that drive? 2nd concern is the fact that my cache is relegated to a single NVMe disk now. Should I move all data on the cache to the array then replace the NVMe drive that's faulty (assuming that it is?) Diags attached. tower-diagnostics-20231024-2016.zip

I guess I need some help understanding VM's. I have a few running, and if I edit the configuration, each one has a Primary vDisk Location which points to a single file. Is this file the entire VM setup, including any changes that have been made while accessing that VM and its features? For example, if I back up the .qcow2 file in the /mnt/user/domains/home_assistant/ path, what am I actually backing up, and what am I not?

I'm a bit confused about permissions. I'm using a Photosync app on my phone to automatically transfer images to an "inbox folder" share on Unraid using FTP. ProFTPd is running as a network service. In the users, I've created a user called `ftpuser` with Read/Write permissions. When I initiate an upload, the files are written to the share as `ftpuser` as the owner and RW only for the user and not for any other groups. When I use Lightroom on my desktop to import those media files, which moves them from the "inbox folder" to a different organizational structure, Lightroom gives me a permissions error. Lightroom is using the share based upon MY user name which I used to connect to the share via NFS. What's the best way to ensure that everyone has full permissions on this folder regardless of which user is adding files to the folder?

Wait wait...I think I got it. I didn't realize you could choose a version using a tag when pulling from the repo. Got it!