Posts by cubemin

    You can also schedule a system backup (from the OMV system partition to a shared folder) with the backup plugin in omv-extras. It allows you to choose from options such as dd, borgbackup, and fsarchiver.


    The third option is making a full image via CloneZilla. Either reboot the server and run CloneZilla from a USB stick, or let OMV itself install the CloneZilla ISO on the system and reboot directly into it. (The next reboot will boot back into OMV.)

    The filesystem that lives on your WSRAID (/dev/md0) isn't mounted for some reason.


    Is the RAID array itself up and working? (Check both Disks and Multiple Devices under the Storage tab.)


    Unless there are other issues, I would think that simply mounting the filesystem from the GUI should make everything work properly. But I have a feeling you would have thought of that already...

    Your data will be safe if you run the fix7to8upgrade script. (That said, I hope you already have backups and aren't just now thinking of making them...!)


    Are you unable to log into the GUI? Quotas can be disabled from the CLI but that's more involved and a bit risky, as it would involve editing OMV's config file.

    I have no idea about the BIOS errors at the beginning of the boot log; those don't look good to me...


    As for the quota service failure, this is easy to fix. Install the mounteditor plugin under omv-extras. It will list all your mounted filesystems and provide buttons to enable or disable disk quotas. Disable them for each filesystem and accept changes.


    After a reboot, the quota errors should be gone. Hopefully that takes care of the long boot time.


    Consider also running the fix7to8upgrade script. It may not be necessary but won't do any harm, and it may catch and fix some unresolved update issues.

    The last message (failed to fetch...) just means that packages.openmediavault.org was temporarily unreachable or you had a brief connectivity problem that resolved itself. Should be harmless.


    The other posts of yours look like an Armbian repository problem. I don't see anything obviously wrong with the corresponding sources.list.d entry. Is this still happening?

    I'm not sure what the group_add directive does in the Compose file, exactly - but I keep seeing PUID 1000 and PGID 100. That should be the first user created by you in the OMV GUI, if I'm not mistaken. It's not root, so my question is who is user 1000? Is this user in the "render" and "video" groups?


    I would also uncomment the two /dev/dri lines (especially renderD128) and see if that helps.

    Off the top of my head I don't remember exactly what is required for VAAPI support inside a container...


    ...but in the case of Emby which I use, I remember that the user the container is running as needs to be a member of the render group and some other video-related group, and that some rendering device under /dev needs to be passed through via a volume definition.


    You could check what HomeAssistant does to be able to use VAAPI (i.e., examine its Compose file).


    Übrigens glaube ich, wir machen lieber auf Englisch weiter, damit andere diesen Thread auch mitverfolgen und u.U. behilflich sein können. 🙂

    You're correct; the container's internal /storage path is bound to the external path you define through the volume directive.

    But the /sharedfolders/Media/obs path is just meant to be a placeholder, an example - you want to replace that path with your own, not add to it.


    Do you speak German by any chance? Asking because the comments in the compose file explain how to change the default settings in the Compose file. (I'm a native German.) I'm just wondering if you're able to understand them or not...

    Well, the container keeps crashing and being restarted (segmentation fault). There's not much detail as to what exactly is going wrong.


    As for your Compose file:

    1. You can remove the version: '3.8' line at the top, it's unnecessary and obsolete.
    2. Did you really mean to bind path /srv/dev-disk-by-uuid-93432eac-e6c2-455e-a3b5-01fd120d98b5/Record/sharedfolders/Media/NDI-Captures to the container's /storage folder? I feel like you just pasted your own path in front of the existing /sharedfolders/Media/NDI-Captures when you actually just meant to use /srv/dev-disk-by-uuid-93432eac-e6c2-455e-a3b5-01fd120d98b5/Record.
    3. Personally, I would also try commenting out the network_mode: "host" line, as well as the PUID and PGID as the container may not work correctly unless run with root permissions.

    You did a lot of troubleshooting before posting this - kudos. I was going to suggest swapping out the SATA cable.


    Unfortunately, everything points to one of the disks failing. Definitely look at the SMART values - you'll probably have rather high numbers for Offline_Uncorrectable and/or UDMA_CRC_Error_Count.


    (Go to Services -> SMART -> Devices, then click on Details for both of your drives.)


    EDIT: On second thought, how are the drives powered? I googled your Fujitsu and I'm getting a vague feeling that the power supply is not up to running both drives. Have you or can you look into that?

    Phew! Thanks ryecoaaron and votdev - that did it.


    Funny story: after issuing omv-mkworkbench all and refreshing the GUI, I was momentarily disheartened to see that Symlinks was still there.

    Then I facepalmed as I realized I'd run the command on my primary server whereas the issue was on the secondary one. Opened Putty to the correct terminal and now all is well. ^^

    I realized that on my backup NAS, I didn't need the Symlinks plugin so I uninstalled it (after removing the one existing symlink I had).


    The 'Symlinks' entry is still showing up under Services in the GUI, and selecting it produces an unsurprising RPC service 'Symlinks' not found error.


    So the actual plugin is gone, but the GUI tab won't go away. I re-installed and uninstalled again, but that didn't help. How to fix?