ZerkerEOD

I disabled it after and it didn't do anything. I found another topic that told me how to stop mover. I already reenabled it, but I was able to start the mover on a single share. So I am going to manually start the mover from the share menu until I find which one was causing issue. Can the System share be moved or does that require any special methods?

I am at my wits end. I have been trying to run mover for a couple hours. I have disabled VM's and Docker and it wasn't moving stuff from the cache to the array. I upgraded to 6.12.3 and the same issues persists. I thought it could be because some of the shares have NFS attached so I disabled them. I rebooted a couple times. I also enabled the logging through scheduler (and unless I am missing other logs) there are no errors. I made all my shares yes:cache or yes:vmstorage trying to move them off of them so that I can swap out the NVMe drives with new ones, but they won't move. Just to add with the new version under share it sows everything should move to the array but nothing is moving. I see where mover starts and every hour mover tries to start again and says that it is already running but it isn't. All my drives spin down after so long and I have forced them back on a couple times. Diagnostics attached and I hope someone can help out. nasty-diagnostics-20230715-1924.zip

It was the docker container or some process with the container. I removed the container completely and brought it back and it did it. I could change my permissions to 777 owner nobody:users and it would stay that was for hours until I manually started the postgresql container, then it instantly failed and the permissions were changed and the owner was unknown and the docker image refused to fully start up and would crash and burn. Looks like I found some other issue that the postgresql appdata CANNOT run from a user share even if it is set to only use a specific drive. Changed the mount point from /mnt/user/appdata/postgresql to /mnt/vmstorage/appdata/postgresql and it booted right up. Something about how unraid handles the file system and postgresql docker image not playing nice. Not sure why it ran for months without issue and then nothing but issues, but oh well.

Did you get this fixed by chance? This is happening to me every time my container tries to start and then it crashes.

I am running into an issue, out of no where my Postgresql container has started changing my appdata/postgresql directory owner to UNKNOWN and permissions of 600. It happens every time I try to run the system. I can change the owership and permissions to 777, but as soon as the docker tries to start it resets everything back to UNKNOWN 600. Any help would be great. I tried adding the PUID and PGID to 99 and 100. The docker won't even start so I can't poke around in that.

I am struggling here. I am trying to change from transmissionvpn to delugevpn. On transmission I can reach all my ports when I have the --net=container:tranmsionvpn. But when I change them to use the deluge client, and restart the stack, the containers have internet, but I can't reach their management ports anymore.

Okay, I did some more digging. It looks like the file permissions change when I start the Postgresql container. I am not sure why this is happening. It is setup to not update and sit at version 14. All of a sudden it just started doing it. Running it as root doesn't help adding the PUID and PGID of 99 and 100 doesn't fix it either. There doesn't seem to be any logs I can find that show it changing in docker logs.

Okay, this happened again today. Here are the diagnostic logs. I hope someone can really help me out. nasty-diagnostics-20230612-2344.zip

I am struggling. I have a container setup for authentik that relies on a postgresql14 container. It was working flawlessly until recently. Over the last couple weeks, I get notified that authentik cannot reach the database, so I go looking, and the appdata share looks good, but the postgresql14 appdata directory is now owned by UNKNOWN with only owner rwe privs. The postgresql instance will not start because it is unable to read the files. So, I have been having to go and manually alter the ownership, and privs and it runs fine for a day or two then bam, it died again. I thought it could have something to do with my backup (using CA backup for docker container). So, I exempted the postgres container from this process and it is still happening. I just uninstalled it as I have switched to BTRFS snapshots for backup, I am hoping that fixes the issue. I just restarted the entire system, so I am not sure what the logs will have. If it does it again, I will try to pull the system logs shortly after getting the notice. Has anyone else had this happened, and if so how did you fix it even if it was something else.

Awesome, thanks for the tips. That worked. Out of curiosity, if I took a snapshop while a VM was running would that cause any issues if I needed to recover from the snapshot?

Can I not make an existing folder a subvolume?

Nothing, it was the directory I was putting the snapshots in. So I feel I am messing up. These are my current directories on vmstorage. I am wanting to get rid of appdata-backup in favor of the snapshots. Currently, I have the script that stops all containers, then copies the appdata directory into the appdata-backup and then restarts them all. But I read somewhere that taking a snapshot is better and you can leave the dockers running. So right now, I just want to snapshot the appdata directory.

I cannot get this working. Every time I take a snapshot the folder is empty. And here is the directory: What I am trying to do, I moved my appdata for the dockers to this drive that is BTRFS, I would like to take a snapshot of it and then use duplicati to backup the snapshot online.

Is there a way to upload a lot of icons to select from instead of individual?

That is a lot of good information and will help me out. But at the time I was seeing the issue, none of my shares were set to yes: prefer was the domain. I will set appdata to my prefer my cache, domain will be on vmstorage and set to preferred, will set system to prefer cache. The anonymized one is probably my data directory that I have all my storage for my old Dockers (not appdata) that contains my media and storage files. One thing I would question is if system should be on a fast pool, why is it default to no? As of right now everything moved correctly to where I had it placed. Not that my initial data migration is complete, I've added the parity drive and running the initial parity. Edit - for the initial data migration I was reading about having everything go to cache then let mover deal with it to speed up. So I did take everything off cache for more room. I think some of my issues were that my initial cache was 500gb and filling way before the hour. After I got the mover plugin and set it with vmstorage and also added to start it at 75% made it way better.

Here are the diagnosis. nasty-diagnostics-20220821-1829.zip

So it's working now after I downloaded CA Mover Tuning. Not sure why but everything is being moved from cache to the array. Would love is someone knows why it was working on reverse without that plugin.

You can ignore my noobness into Unraid and the settings. I found that I have to go to settings then nerdpack and can install the packages I need. I looked under plugins and clicked nerdpack and didn't see anything, so that was my bad.

I have nerdpack installed. Not seeing an additional option to install screen, I thought everything was included in the plugin.

So this isn't working for me. I'm trying to use screen or tmux so that I don't have to migrate my data while keeping the terminal open. I downloaded Krusader and it's horrible, keeps failing telling me it can't write after about 100 gigs. All going to the same share.

Hello I'm having an issue moving files from cache to the array. I have mover set to hourly, and it was working great with my data migration until it didn't. After it wrote the initial amount to the array, I had 330gb on the array. Now when it ran I have 178gb on my cache selected and 256 on the array. I have the share setup with Cache: Yes Device: vmstorage which is my larger ssd until I move all then my cache will be a 500gb and the 1tb will go to VM storage. My understanding is that yes prefer will keep it on the cache whole yes will initially write to it then let mover move it to the array.

That's what I'm thinking, I'm looking into trying to get something and building something from scratch.

How can I pull it from the server then? I just went to the syslog section and copied the text to a file. On a side note, I don't think it's Unraid at all. I downloaded memtest86 booted it up and started running it. At 1 min 20 seconds it froze just like Unraid did. I'm not sure why ADM (Asustor) works perfectly to include large transfer's, but everything else fails miserably. I can't wait to get rid of this POS hardware.

So I have it running (Have to boot through the boot with file option each time, otherwise it tries to boot into ADM). I am about to pull the CMOS battery to see if that fixes the BIOS. However, now that I have it running I started the syslog server after the first crash. I now have a display connected directly to the NAS and using the terminal for the file transfer. I will upload two photos showing that but not sure how it helps. The second photo, with only one line out output from rsync, was taken after it froze. I let it run over night until the morning. The display stayed on and didn't time out like normal, and key presses did not register. I figured maybe its still running in the background or something. I let it run over night and would think that 8TB would transfer in 15 hours. Today's login showed that it wasn't running in the background as I have like 5GB more on the drives, so it failed quickly when starting rsync. syslog.txt nasty-diagnostics-20220807-0923.zip

Does not appear to be the flashdrive. I just got a Samsung fit which is a recommended one and it is also not being recognized at a boot device. I am not sure why it initially worked and now it doesn't.