Beiträge von pavelkms

  • gvt guest page write error

    I've been going crazy for the last few days. My server started to fall off with an error for no reason. Everything seems to be working as it should, then bang and the server is offline, only a reboot via reset helps. And you won't look at the logs, because they are in RAM and after such a reboot they are not saved. The error was not regular, it may work normally for a day, or it may fall off an hour after the reboot. Fortunately, there is PiKvm, after the dump I connect and see a bunch of incomprehensible errors on the screen, each time different. Periodically I saw familiar letters - PCI, SATA, BUS in different variations, with different numbers and the word error. The drives of all the disks are in order, I changed the cables, nothing helped. When studying logs after dumps on the screen, PiKvm noticed that the phrase - gvt guest page write error periodically comes across. I Googled it, it turns out that this is a known error when using gvt when running an intel video core into virtual machines. Moreover, it occurs rarely and not at all. To occur, it is necessary that the hardware configuration, software and Linux kernel coincide unsuccessfully, not every kernel happens, in general, something abstruse. Everything seems to be working, errors are slowly accumulating, then bam and the system crashes. There are some patches, different for different cores, in general, a murky story. I have been using gvt for a long time, I have never had such problems. I turned off the video core in the Windows VM and did not make a single error all day. Then I changed the stock core in OMV to the proxmox core, returned the video to the VM and everything worked fine all day.

    In general, the pve core did not help much, the gvt guest page write error still appears, although much less frequently. If everything used to fall off abruptly, now it can just start to blunt some container. For example, yesterday everything worked well all day, my wife just called from work in the afternoon, complained that Nextcloud was working slowly, looked at it, it seemed fine, they decided that her Internet was bad. In the evening, I looked into the Nextcloud logs, and there his database is full of errors, I tried restart, half of the AIO containers stupidly refuse to start. I started to reboot and then the gvt guest page write error came down. Before the reboot, all other containers and VMs were working fine all day. Although no, on a Windows VM at work during the day, sometimes it was thrown out of the RDP session for no reason. Reconnected, and everything seems to be fine, I also thought about Internet glitches. I have removed the video core in the VM for now, we will continue to look.

  • What Features to come with OMV 7 (What do people hope for/expect)?

    Zitat von ryecoaaron

    The web interface is a responsive design. You can use it on a mobile device. I'm not sure who would write an app otherwise.


    The first can be done using the current OMV web interface on a mobile device. The latter would need a change to the plugin and since I don't use the plugin, I don't have the familiarity to add that feature.

    For the latter, I now use Telegram, very convenient.

    https://4pda.to/forum/index.php?showtopic=1065136&view=findpost&p=124448721

  • KVM Plugin feature request

    Hello!

    I want to forward the VNC port in my router and connect to my Debian 12 VM over the internet. How can I secure vnc connection and add permanent password for VNC connection?

    I edit the xml, paste it there - passwd='12345', in the graphics section:

    Code
        <graphics type='vnc' port='-1' autoport='yes' listen='0.0.0.0' passwd='12345'>
      <listen type='address' address='0.0.0.0'/>
    </graphics>


    When I start the VM, connect to the VNC, I am asked to enter a password, I enter 12345, and I get into my VM, everything is fine.

    But on the next restart, my passvd='12345' disappears and the xml returns to its original form.

    How to make it permanent?

  • How to free port 25 for mail server?

    I can't show you any documentation about port 25. But for a long time, in the manner described above, I replaced port 25 with port 26 in OMV and everything works for me. OMV can send its messages without problems and my mail server works with port 25.

  • How to free port 25 for mail server?

    Zitat von ryecoaaron

    Why do you not want to use macvlan? OMV depends on postfix and OMV controls the postfix config. Trying to hack OMV to not use port 25 is a LOT more effort than just using macvlan.

    I made it very simple:

    nano /srv/salt/omv/deploy/postfix/30master.sls

    Code
    customize_postfix_master:
file.replace:
- name: "/etc/postfix/master.cf"
- pattern: "smtp      inet  n       -       n       -       -       smtpd"
- repl: "26      inet  n       -       n       -       -       smtpd"

    After reboot, postfix in OMV listens to port 26. Port 25 is now free for dockerized mailcow.

  • Hardware upgrade and OMV

    I am not a professional in Linux, can you give instructions on how to properly remove the network bridge, please?!

    Or can I just delete the br0 network device in the OMV GUI and then re-create it in omv-firstaid?

  • Hardware upgrade and OMV

    I tried to add an Nvidia graphics card to my OMV6 and lost network connection to OMV after that. I assume that adding a new pci device changed the name of other devices in the system, omv-firstaid did not help me fix the problem. Maybe because before that I changed my network connection to the br0 bridge. When I removed Nvidia, everything worked as before. I don't know how to add Nvidia to OMV now without reinstalling the whole system.

  • docker not working since omv-upgrade

    Good afternoon!

    Now I installed the apparmor-utils, restarted the docker and everything worked for me.

    In the future, I may have some problems with docker? Maybe it's better for me to remove the apparmor-utils now and disable the apparmor in system?