Posts by Olivier

    This is expected and as designed. resolvectl should not be broken on a properly working OMV system. So, it is difficult to avoid that timeout if resolvectl is broken.

    100% agree the module now works even with a goofy network config.

    DNS issues can also come from my router which is known for its "specific" behaviour.

    I am no expert in Network configs but I will dig on the subject using this post:

    systemd-resolved not resolving any domains
    My NetworkManager connects to the internet but I do not have any internet access due to a bad DNS configuration. I do not have much experience with DNS…
    askubuntu.com

    Here is the content of my resolv.conf:



    Once the timeout is reached I have 2 items in the list:

    • "disabled"
    • "custom" wich displays a new field to specify a DNS manually

    I somehow managed to make the Wireguard tunnel work for a few minutes when I replaced my duck DNS name on the tunnel with my public IP. I created the tunnel, the client on OMV from scratch and restarted the client (phone) and now it works.

    I suspected now either the phone or DuckDNS to be the culprit.

    I agree but I wanted to see the output from your system first.

    Jeez dude. The plugin didn't handle a system in an abnormal state. I haven't seen this on any of my systems. I can't think of everything... Not like I leave plugin purposely in a broken state. I will fix it but just have some patience.

    No problemo man, I can't blame you for that!

    I'm happy to give you a detailed feedback so you can make this plugin more robust.

    I think in case theresolvectl status command fails the pluging should at least populate the "DNS Servers" drop down with the "Disabled" value.

    The pluging can trace this with a warning message.

    No offense but for me it is a regression from the previous version that need to be fixed.

    Here are my outputs:





    All the services were up and running.


    I managed to fix this issue TEMPORARILY by configuring the FallbackDNS parameter in file /etc/systemd/resolved.conf


    Code
    FallbackDNS=127.0.0.1 ::1


    After that I could get all the available DNS Servers using theresolvectl status command :


    This post helped me :

    systemd-resolved - ArchWiki


    The dropdown is now populated but... /etc/systemd/resolved.conf is overwritten once I restart the system

    It seems that avahi is managing this part and the config of this tool is generated by openmediavault.

    But the DNS server is properly configured in Network->Interfaces (enp1s0)

    It's as if the config was not applied to resolved.conf

    Hi


    I was using the WireGuard successfuly using for a very long time.

    I updated the OMV plugin recently and couldn't use it anymore.


    I struggled to find out the source of the pb and I noticed a new dropdown list on the client config named "DNS Servers"

    The list is empty and when I click on it to select an item, a neverending loading icon is displayed.


    I noticed errors poping from the notification bell:


    I tried to run the faulty command manually and I got this error:


    Code
    root@omv:~# export PATH=/bin:/sbin:/usr/bin:/usr/sbin:/usr/local/bin:/usr/local/sbin; export LANG=C.UTF-8; export LANGUAGE=; resolvectl status | grep -oE '\b([0-9]{1,3}\.){3}[0-9]{1,3}\b'
    Failed to get global data: Connection timed out
    root@omv:~# resolvectl status
    Failed to get global data: Connection timed out


    Any idea ?



    Thanks in advance

    I had one of the HDD off in my RAID 1 for some times, and when I managed to power it on again I could not find the resync/recovery status.

    Once the openmediavault-md module was installed I was able to see the recovery status !


    Thanks !

    I measured power consumption from the brick plug.

    The NAS takes max 39 watt AT startup and then arround 12-19 watt AT cruise speed.

    2 years ago I had this startup issue. I made a little script to reboot the system if a mount point IS missing. That dirty fix worked rather well, in fact the missing HDD even gradually booted corretly again.


    But we had some storms and lightnings recently and the problem went black since then.

    I suspect micro power cut outs to be the main cause of this recurrent issue.


    EDIT: I finally identified the culprit! I googled and found this old post:


    In fact the J3455 is very low profile. The main issue comes from the fact that the motherboard do not drains enought power.

    I used the recomandations of the post :

    - by reducing CPU state to C6

    - by setting the CPU Mode to Sport

    - by speeding up the case fan a little to ensure a minimal consumption.

    The last point may seems a goofy trick but ensures an steady average consumption


    After that all my HDD were detected and worked as expected!


    ADDENDUM:

    Disable "Aggressive SATA Device Sleep" feature as it is known to case HDD issues

    Hi Ixox,


    Nice build, very neat enclosure.

    I have almost the same config : ITX-J3455 motherboard, 8 go RAM, 90 watt Pico PSU.

    Main difference I made it with 4 bays drawer style in front.

    I used to recycle big HDD with a good capacity and using only 2 at the same time.

    Recently I replaced one of them with a Raid 1 (2 x) IronWolf 4To.

    Since this upgrade I had many problems at startup, one of the drive seems to remain undetected.


    I suspect a power shortage, but I don't now if I should upgrade the PSU wattage.

    I have two options:

    • change for a pico PSU. That means changing both PSU unit AND the power adapter (~70€)
    • go for a flex PSU (different from pico PSU scale). It is cheaper (~35€) but with a fan it can be noisier, and the adapter would be located inside the case inducing more heat into HDD enclosure

    Which solution should I choose ? and with which wattage ?


    thanks in advance

    Ok that's legit.

    Don't bother spending time on a deprecated feature.

    That was a "nice to have" request but if AFP module is no more available I'm OK with that.

    Moreover I recenty tuned up the SMB service using this post and it seems to be slightly better:

    Open Media Vault SMB performance quick win
    A quick performance win on transferring large files with Samba. Even if you use Open Media Vault or not, this might help you. Give it a shot! Here's what to
    techie-show.com

    Hi there,


    I started to use Photoprism several years ago and I'm very pleased with this photo library.

    I struggled to make it work with the intergrated plugin so I went for the Docker Compose solution.


    I found it highly configurable to match my needs and also the resources of my hardware.

    I can tune up CPU usage, RAM usage, face recognition frequency, default Theme, password and many other options.

    The only thing I am missing is the HTTPS access using OMV certificates, but as I use it with local access only it is not a big issue for me.


    I recently tried the plugin option again and managed to make it work at last.

    The pluggin pulled many dependencies like caddy to activate https and many other different things.

    In addition to that the pluggin expose very few parameters: photo folders, certificate to use, and that's all. No CPU or memory max alloc settings.


    All that stuff packed together made my system run as slow as a cow.

    I saddly had to go back to the Docker solution, wich was a less straight-forward but a much more efficient setup.


    Volker Theile is in charge of this plugin.

    Do yo think there is a chance he can add some extra parameters ?

    Objectively AFP is much faster and much more reliable.

    Since I switched to SMB I faced many issues which never happened with AFP.

    I don't understand why the module was remove, it should be available at least in OMV extra.

    Not to mention the OMV AFP module was very convenient to setup shared folders.

    Hi there,


    I have more or less the same kind of setup.

    I use the Rsync module with scheduled tasks.

    - on iMac account I created a SSH key

    - on OMV I registered the SSH key for each account so OMV can access them

    - I have one 'PULL' scheduled task for each account

    - I activated daily backup at 01:00 AM and email notifications to be sure the backup is performed


    The main advantage is that the machine+account you backup only has to provide SSH access.

    And that's all, not backup task to setup on the target, everthing is centralized on OMV.

    If your docker container is properly configured for SSH, you should be able to use the rsync 'PULL'


    How to setup password-less SSH access: https://phoenixnap.com/kb/setup-passwordless-ssh

    Rsync 'PULL' mode: https://forum.openmediavault.org/index.php?thread/24751-rsync-pull-mode

    Hello


    I had a similar issue with my NAS based on a J3455 motherboard with 4go of regular DDR3 ram.

    I tried a lot of things but nothing fixed the problem.

    I recently upgraded the RAM for a set of 2x8Go DDR3L low-profile optimized to run with 1.35v and well supported by the J3455.. and since then everything works fine!


    It solved my issue.

    I thinks most of the time it is link to power consumption.


    I hope this will help!

    Ok I found a rather "clean" way to handle it:


    1-Install the APT-tool available in OMV plugins

    2-Go in Services > APT Tool > packages

    3-Add package named "software-properties-common" and install it

    4-Connect with ssh or open a terminal on the OMV server

    5-Add you custom repository using the command:

    Code
    sudo add-apt-repository ppa:<repository_name>
    sudo apt update 

    The "apt update" command might return a message like:

    Code
    W: GPG error: http://ppa.launchpad.net/<repo_name> <version> InRelease: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY 0000000000000000
    
    E: The repository 'http://ppa.launchpad.net/<repo_name> <version> InRelease' is not signed.


    Then you just need to add the public key of the package to your trusted list using this command:

    Code
    sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-key 0000000000000000


    Then you can run the apt update command again and install the packages from your custom repo.



    I hope this will be of some help !


    Regards