Posts by hoppel118

  • Bridge and VLAN

    Hi,


    I also wanted to do this. But got the following message with omv6:


    Quote

    500 - Internal Server Error
    Failed to execute command 'export PATH=/bin:/sbin:/usr/bin:/usr/sbin:/usr/local/bin:/usr/local/sbin; export LANG=C.UTF-8; export LANGUAGE=; omv-salt deploy run --no-color systemd-networkd 2>&1' with exit code '1': debian: ---------- ID: remove_systemd_networkd_config_files Function: module.run Result: True Comment: file.find: [] Started: 16:08:15.368985 Duration: 1.238 ms Changes: ---------- file.find: ---------- ID: remove_empty_systemd_networkd_config_files Function: module.run Result: True Comment: file.find: [] Started: 16:08:15.370332 Duration: 0.89 ms Changes: ---------- file.find: ---------- ID: remove_netplan_config_files Function: module.run Result: True Comment: file.find: ['/etc/netplan/10-openmediavault-default.yaml', '/etc/netplan/20-openmediavault-eno1.yaml', '/etc/netplan/60-openmediavault-br0.yaml'] Started: 16:08:15.371312 Duration: 1.168 ms Changes: ---------- file.find: - /etc/netplan/10-openmediavault-default.yaml - /etc/netplan/20-openmediavault-eno1.yaml - /etc/netplan/60-openmediavault-br0.yaml ---------- ID: configure_etc_network_interfaces Function: file.managed Name: /etc/network/interfaces Result: True Comment: File /etc/network/interfaces is in the correct state

    Did you ever solve your issue?


    Thanks Hoppel

  • Auto-Spindown of hdd's does not work on OMV6

    I didn't need to install gnupg2. But, yes true words. tinkering with hdparm seems to be pointless at the moment.


    I also had spindown issues with my 8 WD Red 10TB disks. In my case not all drives made the spindown, here an example:


    Sometimes I saw 7 of my 8 disks in standby. Why not all disks? I don't understand this.


    The following command always brought all 8 disks to standby:


    Code
    hdparm -y /dev/sd[abcdefgh]


    My hdparm.conf was configured through OMV6 web interface and looked as follows for all eight disks:


    Code
    /dev/disk/by-id/... { 
apm = 127 
spindown_time = 120 
write_cache = off
}

    Played around with the hdparm settings by the omv web interface, but it didn't get better.


    So I installed hd-idle and it seems to do what is expected. Thanks for the link to the good How-To.


    Regards Hoppel

  • New openmediavault-kvm plugin

    Quote from ryecoaaron

    What I suggest doing and what you did are very different things. I guarantee installing an updated osinfo-db package and NOTHING else will not cause a problem. I guess I could just put the package in the omv-extras repo.

    😃🤙


    omv-extras repo would be a good workaround for me. Can you also make this package a dependency for openmediavault-kvm?


    Good night… 😂

  • New openmediavault-kvm plugin

    Quote from ryecoaaron

    It is not sorted in complete alphabetical order. Look right before Mac OS. The plugin is just return the output of the command and does not determine the sorting order.

    Ok, only „generic linux“ and „microsoft“ are not in the right position…


    But yes, you are right it’s there. 😃

  • New openmediavault-kvm plugin

    Quote from ryecoaaron

    It makes very little difference which one you pick

    Ok, this what I wanted to „hear“.


    Quote from ryecoaaron

    but while Generic Default doesn't exist, Generic Linux 2016, Generic Linux 2018, Generic Linux 2020 do exist.

    I don’t see this „Generic Linux 20xx“ options:



    Quote from ryecoaaron

    This list is populated by the osinfo-query command which gets its data from the osinfo-db package. If you want to update it, use the package from bookworm (I use the updated package on all of my servers):

    Thanks for this hint. But I am unsure, if I really want to install packages manually after my omv4 situation… 😉

  • New openmediavault-kvm plugin

    Quote from STUKguy

    I am assuming i download the zip for .qcow2 and extract to the the folder on the OMV server

    Yes, you have to unzip the file to use it. At least this is what I did.


    Who is the owner of the qcow2 file?

    Code
    ls -l /data/vm/haos_ova-9.4.qcow2
-rw-r--r-- 1 libvirt-qemu libvirt-qemu 7561936896 24. Dez 00:28 /data/vm/haos_ova-9.4.qcow2

    Is it libvirt-qemu?


    Can you describe (maybe with screenshots) how you created your pools? On which filesystem is your pool? Which architecture is your server?

  • KVM Plugin - Remove USB Device from VM not possible

    Hm...


    So I did the following:


    • stop the vm
    • deleted the following part of the vm.xml
    Code
        <hostdev mode='subsystem' type='usb' managed='yes'>
      <source>
        <vendor id='0x10c4'/>
        <product id='0xea60'/>
      </source>
      <address type='usb' bus='0' port='4'/>
    </hostdev>

    --> It doesn't work! Still the same issue:


    Code
    500 - Internal Server Error
Failed to add USB device. error: Failed to attach device from /tmp/virsh_usbIsJpae error: Angeforderte Operation ist nicht gültig: Einheit ist bereits in der Domain Konfiguration vorhanden virsh attach-device --domain haos --file '/tmp/virsh_usbIsJpae' --persistent --config --live
    Code
    cat /tmp/virsh_usbIsJpae 

<hostdev mode="subsystem" type="usb" managed="yes"><source><vendor id="0x10c4"/><product id="0xea60"/></source></hostdev>


    So, I decided to try the following:

    • stop the vm
    • deleted the snapshots of the vm
    • ls /etc/libvirt/qemu/ the file vm.xml is not there anymore
    • recreated my vm with the same qcow2 file
    • added the usb stick -> YEAH!!!! THIS WORKS!!!! :)
    • started the vm -> Yes, it works!

    Anyway, thanks! ;)

  • KVM Plugin - Remove USB Device from VM not possible

    Hi, I have the same issue.


    Yesterday I setup a VM with Home Assistant Operating System. So far so good. Now I attached my Home Assistant SkyConnect USB Stick (ZigBee/Thread Connectivity) via the KVM Plugin and it was recognized in the VM.


    But I am absolutely new to Home Assistant. So I worked with Snapshots. After some misconfiguration I went the route to go back to an earlier snapshot of that vm. But I had another issue and decided to delete the vm including its snapshots and recreated the vm with the qcow2 file.


    But now its not possible to add the Home Assistant SkyConnect again. I see messages like the following in a red box:


    Code
    500 - Internal Server Error
Failed to add USB device. error: Failed to attach device from /tmp/virsh_usbXQcqvY error: Angeforderte Operation ist nicht gültig: Einheit ist bereits in der Domain Konfiguration vorhanden virsh attach-device --domain haos --file '/tmp/virsh_usbXQcqvY' --persistent --config --live



    Code
    root@omv6:~# cat /tmp/virsh_usbXQcqvY

<hostdev mode="subsystem" type="usb" managed="yes"><source><vendor id="0x10c4"/><product id="0xea60"/></source></hostdev>

    It is this devices:


    Code
    lsusb
...
Bus 001 Device 004: ID 10c4:ea60 Silicon Labs CP210x UART Bridge
...

    My vm.xml looks as follows:



    What exactly do I have to change or delete in the vm.xml file? The relevant part seems to be this:

    Code
        <hostdev mode='subsystem' type='usb' managed='yes'>
      <source>
        <vendor id='0x10c4'/>
        <product id='0xea60'/>
      </source>
      <address type='usb' bus='0' port='4'/>
    </hostdev>

    Is it as simple as removing the complete part or only parts of it?


    Thanks Hoppel

  • New openmediavault-kvm plugin

    Hi STUKguy


    I did this yesterday. I decided to setup Home Assistant Operating System in KVM.


    Here is a good general KVM HowTo: [How-To] Use the Openmediavault-KVM plugin


    • install openmediavault-kvm
    • create two folders (pool and isos) -> chapter 3.1
    • configure pool and isos as storage group -> chapter 3.2
    • network configuration -> chapter 3.3 (If you want to do it right, you want a bridge interface in OMV)
    Quote

    If we want a network interface that has an IP in our LAN and at the same time can access the shared resources of the host, we must create a bridge interface in OMV (in the Network>Interfaces section) as follows. Warning: This method will remove the current network interface configuration and replace it with a jumper. If you have services configured on the host with this network interface (for example, Wireguard), you will need to reconfigure them to work with the bridge:

    • Go to Network>Interfaces> Make a note of the network interface name that appears under the Device column. Select this interface and click delete, BUT do not apply changes yet.
    • Without applying the changes of the previous point, click on + Create then on Bridge. In the Interfaces field, find the name of our newly removed interface and select it. Enter the necessary data for the Ipv4 protocol with static IP, for this in Method select Static. Fill in the fields:
      • IP: Your OMV IP
      • Netmask: 255.255.255.0
      • Gateway: IP of your DHCP server (your router)
      • DNS: For example 1.1.1.1
    • Press Save and now you can apply the changes. Now your OMV network interface will have a name similar to br0. This is the bridge-type interface that you should select on your VM.
    • Add the newly created network interface (br0) to your VM. You should not create an interface in KVM, macvtap or any other.
      • If you already have a VM created:
        • stop it, remove the current interface assigned and simply select your VM and click on Network and then Add Network. In Model change virtio to bridge. In Bridge look for the network interface created in OMV, it will be br0 (or similar). Press Add.
      • If you are creating a new VM:
        • In the VM creation window, go to the Network section and change virtio to bridge. In the next field select your network interface br0.


    • configure the access to the vm -> chapter 3.5

    This is how it works for me.


    ryecoaaron According to the HA documentation, we should select "Generic Default" for the operating system. However, this does not exist in the list. I decided to try Debian 10 and it worked. Do you think there is anything wrong with this? Is it possible to add "Generic Default" to the list? Debian 11 is also missing.

  • Information about ZFS cache and hitrate

    Mostly movies and private pics/videos. There are no constant accesses to the raid-z2 pool. My docker containers and my virtual machines have its own SSD (without redudancy). But I have backups on my zfs storage. May I am going to change this behavior in the future and buy two new SSDs, which I use in a zfs mirror for my docker/vm pools/storage.


    For the moment I don't see any reason to change something. ;)


    Thank you

  • Information about ZFS cache and hitrate

    Hi,


    so what does "ZFS Hits/Misses" mean?


    My Widget shows the following ZFS Hits/Misses ratio:



    My pool looks as follows:

    (Yes, I know there is a zpool feature upgrade. Can't do this at the moment.)


    Code
    @omv6:~# zpool list
NAME        SIZE  ALLOC   FREE  CKPOINT  EXPANDSZ   FRAG    CAP  DEDUP    HEALTH  ALTROOT
mediatank  72.8T  47.9T  24.8T        -         -     9%    65%  1.00x    ONLINE  -

    I don't have an L2ARC cache device:


    Code
    @omv6:~# arc_summary -s l2arc

------------------------------------------------------------------------
ZFS Subsystem Report                            Thu Dec 22 13:42:34 2022
Linux 6.1.0-1-pve                                             2.1.7-pve1
Machine: omv6 (x86_64)                                        2.1.7-pve1

L2ARC not detected, skipping section

    I have 64GB RAM:



    Maybe I should increase the maximum amount of memory that ZFS can use for ARC caching.


    I don't have and never had any issues with my zfs. But seemingly I have some spare RAM. :D


    What do you think?


    Thanks Hoppel

  • Proxmox Kernel

    Quote from crashtest

    With that said, you can keep the older pool as long as you like. So far, newer versions of ZFS have been backward compatible to older pools.

    Yes, I know. But thanks for the explanation and the confirmation. I will wait some days to be sure that everything works as expected. I already did such a feature upgrades while migrating from OMV3 to OMV4. :D


    Today I migrated from OMV4 to OMV6. That was not so easy, but now it works. More details: here


    This piece of software looks and feels really good! Seems to be good proxmox competitor nowadays. :D


    Thank you guys!

  • Boot OMV4 on a physical disk in a kvm guest

    OK, I got it. It works.


    My procedure was as follows:

    1. installed omv6 to new ssd
    2. installed the proxmox kernel
    3. installed openmediavault-zfs
    4. imported my zfs pool / filesystems
    5. configured the users and groups with the same ids as before
    6. configured the nfs and smb shares
    7. configured portainer
    8. migrated my docker containers (via backups)
    9. installed openmediavault-kvm
    10. configured the pools for vms and isos
    11. created a new omv4 vm (downloadeded an old omv4.iso image and configured it completely) -> name and format: omv4.qcow2
    12. converted my physical omv4 to virtual via qemu-img convert -O qcow2 /dev/sda /data/vms/omv4.qcow2 -> same name and format: omv4.qcow2
    13. exchanged the omv4.qcow2 files
    14. deactivate smb/nfs/docker/zfs in the vm
    15. configured the new network interface via noVNC
    16. passed through a homematic usb stick to the vm
    17. Thats it! :D

    Sounds easy, but took me round about 12 hours.


    ryecoaaron Thanks for the following things:

    • the integration of portainer. This was a really smooth migration of my docker containers, even though i just liked the good old omv-docker plugin. ;)
    • the development of the kvm plugin. Now I really don't need the oversized Proxmox anymore. Even LXC is possible. WOW!!!
    • thanks for the development of omv-zfs and providing the proxmox kernel in an easy way. This is the base of all for me.

    votdev and ryecoaaron Thanks for the hard work you investigated over the years. Everything looks and feels really good now. I wish you a Merry Christmas and a happy new year.

  • Proxmox Kernel

    Quote from macom

    To force a reload of the browser page. Might solve your issue.

    Thanks. I rebooted the server without success. However, I did not get the idea to refresh the browser. This key combination doesn't work for me on Macs. For Mac user it is [Command + Shift + R].


    But it worked when I imported the pool on the command line.


    Quote from crashtest

    Yes. Under System, omv-extras, Settings. Uncheck the Backports box and update.
    ______________________________________________


    This is common behavior. ZFS doesn't (always) keep up with backports kernels. ZFS eventually catches up, but when a new backports kernel comes out it may not be supported by ZFS for awhile.

    The safest route is to install the Kernel Plugin and go with the Proxmox Kernel.

    Thanks, found the option and disabled the backports by the official way.


    ZFS seems to work as expected now.


    I am not sure yet when I will upgrade the zpool. My zpool uses the features of the zfs that came with omv4. =O


    Thanks again.

  • Proxmox Kernel

    Hm..., ok the plugin is installed:



    But when I click on "Storage - zfs" I get the following message:



    Any idea?


    EDIT: OK, I got it. After importing the pool on the command line, the zfs gui works now and I see my pool and my filesystems. Is this a planned behavior?


    Thanks Hoppel

  • Proxmox Kernel

    Hi,


    I am setting up a new openmediavault 6 server with zfs.


    This what I already did:


    1. Installed openmediavault 6
    2. installed omv-extras
    3. installed latest proxmox kernel (Linux 6.1.0-1-pve) and booted it
    4. tried to install openmediavault-zfs and get the following messages:



    So I looked at the command line and see the following;



    So, I googled a bit and found that you need to "disable" the backports:


    Code
    @omv6:~# nano /etc/apt/sources.list.d/openmediavault-kernel-backports.list
#deb http://httpredir.debian.org/debian bullseye-backports main contrib non-free

    after apt update && apt upgrade it works:



    Is this a planned behavior? Is what I did the correct way? Is there a settings in the OMV WebUI to disable the backports?


    Thanks Hoppel

  • Boot OMV4 on a physical disk in a kvm guest

    Quote from ryecoaaron

    OMV is Linux and it can handle any kvm config. So, the default will work. You should read the kvm guide.

    I am not completely new to kvm/lxc, but I never migrated p2v. That’s the reason why I asked for ideal vm settings.


    Played a bit with proxmox in the past. But my main usecase is not the virtualization, it’s a network attached storage.


    With your omv-kvm we have kvm and lxc now. No need for proxmox anymore (at least it seems so)… 😃


    Thank you so much. I keep you updated.


    Hoppel