Jump to content

methanoid

Members
  • Posts

    587
  • Joined

  • Last visited

  • Days Won

    1

Posts posted by methanoid

  1. Coming back to unRAID after couple years away and struggling

     

    I am trying to pass my existing Win10 install on NVME to a VM. So I have my RX6800 (and an R9 270x) both stubbed, two USB controllers stubbed, one 500GB NVME stubbed. I've got the UUID adjusted as per SIO video. I have tried with and without GPU rom being included. VM set to Q35 and UEFI as install on NVME was UEFI. 

     

    Basically when I start the VM I get an unRAID logo (I guess the EFI for VMs now added a logo, nice!) which flashes with text in top LHS which says "Reset System". 

     

    Is this FLR Navi reset bug (I thought that was fixed on RDNA2 GPUs)? Or something else?  Sig contains specs etc

     

    EDIT: Checked logs (that helps!!) "2023-07-09T12:31:22.527758Z qemu-system-x86_64: vfio: Cannot reset device 0000:11:00.1, depends on group 34 which is not owned."   seems to suggest there is another USB device I need to pass?  TRIED stubbing it and adding but made no difference, other than to remove the error message so now nothing in logs indicates any error OR warning ?!

  2. On 10/7/2022 at 9:23 PM, SimonF said:

    PPC aarm64 and riscv have been added. But no template support just binaries.

     

    I missed the notification for this but not running unRAID right now but coming back V soon. 

     

    I dont suppose you know if you have SIO's MacInABox running (it adds the Virt-manager interface) if the Qemu-PPC and Qemu-Arm etc options show up in that? I'd love to have my unRAID box (when its back) run everything I could ever want ;-)

  3. On 4/11/2023 at 10:45 AM, binhex said:

    first issue is to remove the now defunct pia name servers, they start with 209.x.x.x, defined via NAME_SERVERS, from your log:-

     

    nameserver 209.222.18.222
    nameserver 84.200.69.80
    nameserver 37.235.1.174
    nameserver 1.1.1.1
    nameserver 209.222.18.218
    nameserver 37.235.1.177
    nameserver 84.200.70.40
    nameserver 1.0.0.1

     

    i see you also have issues loading iptable_mangle module, which means external access to the web ui may not work, try accessing the web ui from inside of your lan, from your log:-

     

    insmod: ERROR: could not load module /lib/modules/iptable_mangle.ko: No such file or directory

     

     

    Thanks for taking time to respond, I do appreciate it. ALL GOOD NOW, I suspect also updating docker engine (to fix an issue with LS.io Sonarr docker) also helped

     

    The iptable_mangle.ko is missing disappeared when I changed to "Latest"

     

    Nameservers came from the YML file that originally set up the docker even though I never used PIA. 

  4. I've been stuck on old version for a while now as it stopped working with recent updates.  Tried today with "latest" and still refuses to bring up web interface

     

    binhex/arch-qbittorrentvpn:4.4.5-1-01  seems to be last working version for me (I'm using Privado not PIA like many here do).  It works but there's always that worry that you need latest versions

  5. 1 hour ago, methanoid said:

     

    Sorry, IDK. It was all installed by Plexguide/PTS scripts and its always showed 3 ports for Qbittorrentvpn (6881,8098,8118) as per the YML I posted earlier.  I'm GUESSING its down to the Traefik not opening the ports to the world? IDK, I didnt change anything but Qbittorrent suddenly stopped responding so I guess it was the update that triggered whatever broke

     

     

    @binhex   well I solved the problem. Any version newer than binhex/arch-qbittorrentvpn:4.4.5-1-01 (6 months old) will not work for me. I made sure I had new OVPN files from Privado (my VPN provider).    So I am working now but no idea why I cannot run latest versions

     

    That old version of Qbit also comes with the `host not found (non-authoritative) try again later` error on the tracker which requires running 
    echo "nameserver 8.8.8.8" > /etc/resolv.conf
    within the docker to get everything working. 

  6. 5 minutes ago, binhex said:

    it could, yes!, also what port is the webui actually running on? you seem to be flip flopping ports here:-

     

     

    is it 6881 or is it 8998? just ensure you arent tripping youself up here.

     

    Sorry, IDK. It was all installed by Plexguide/PTS scripts and its always showed 3 ports for Qbittorrentvpn (6881,8098,8118) as per the YML I posted earlier.  I'm GUESSING its down to the Traefik not opening the ports to the world? IDK, I didnt change anything but Qbittorrent suddenly stopped responding so I guess it was the update that triggered whatever broke

     

  7. 22 minutes ago, binhex said:

    are you sure you have your container set to a fixed ip?, if you are specifying bridge then the ip you are using to curl to should be your host ip not the container ip.

     

    so try http://localhost:<port for qbittorrent webui> via ssh on the machine running docker.

     

    curl -v https://localhost:8998

     

    * Rebuilt URL to: https://localhost:8998/
    *   Trying ::1...
    * TCP_NODELAY set
    * connect to ::1 port 8998 failed: Connection refused
    *   Trying 127.0.0.1...
    * TCP_NODELAY set
    * connect to 127.0.0.1 port 8998 failed: Connection refused
    * Failed to connect to localhost port 8998: Connection refused
    * Closing connection 0
    curl: (7) Failed to connect to localhost port 8998: Connection refused
     

    Same with http also...  (I have Google auth on all containers and Traefik doing proxy if that could be the issue) ??

     

  8. On 2/22/2023 at 12:10 PM, methanoid said:

     

    curl -v 172.18.0.13:6881
    * Rebuilt URL to: 172.18.0.13:6881/
    *   Trying 172.18.0.13...
    * TCP_NODELAY set
    * connect to 172.18.0.13 port 6881 failed: Connection refused
    * Failed to connect to 172.18.0.13 port 6881: Connection refused
    * Closing connection 0
    curl: (7) Failed to connect to 172.18.0.13 port 6881: Connection refused

     

    and

     

    curl -v 172.17.0.14:6881
    * Rebuilt URL to: 172.17.0.14:6881/
    *   Trying 172.17.0.14...
    * TCP_NODELAY set
    * connect to 172.17.0.14 port 6881 failed: Connection refused
    * Failed to connect to 172.17.0.14 port 6881: Connection refused
    * Closing connection 0
    curl: (7) Failed to connect to 172.17.0.14 port 6881: Connection refused
     

     

    @binhex  any ideas what this tells me (I tried with http:// prefix and https:// and no prefix)

  9. 2 hours ago, binhex said:

    ssh onto your remote machine running the docker container and curl the qbittorrent web ui e.g.:-

     

    curl -v http://<ip of container or host>:8098

     

     

    curl -v 172.18.0.13:6881
    * Rebuilt URL to: 172.18.0.13:6881/
    *   Trying 172.18.0.13...
    * TCP_NODELAY set
    * connect to 172.18.0.13 port 6881 failed: Connection refused
    * Failed to connect to 172.18.0.13 port 6881: Connection refused
    * Closing connection 0
    curl: (7) Failed to connect to 172.18.0.13 port 6881: Connection refused

     

    and

     

    curl -v 172.17.0.14:6881
    * Rebuilt URL to: 172.17.0.14:6881/
    *   Trying 172.17.0.14...
    * TCP_NODELAY set
    * connect to 172.17.0.14 port 6881 failed: Connection refused
    * Failed to connect to 172.17.0.14 port 6881: Connection refused
    * Closing connection 0
    curl: (7) Failed to connect to 172.17.0.14 port 6881: Connection refused
     

     

  10. 5 minutes ago, doubleopinter said:

     

    curl -v yourhost:8080

     

    Thanks. I didnt know whether you meant

     

    curl -v qbittorrentvpn.mydomain.com (??)

    curl -v qbittorrentvpn.mydomain.com:8080 (not a QBT port)

    curl -v 172.17.0.14:8098 (my bridge network)

    curl -v 172.18.0.13:8098 (my plexguide network)

     

    What am I looking for too? 

     

    I also tried   curl -v qbittorrentvpn.mydomain.com:8098

    * Rebuilt URL to: qbittorrentvpn.xxxxxxxxx.co.uk:8098/
    *   Trying 2a06:98c1:3120::3...
    * TCP_NODELAY set
    *   Trying nnn.aaa.bbb.ccc...
    * TCP_NODELAY set

     

    The nnn.aaa.bbb.ccc is NOT my current IP address which makes me wonder if Cloudflare DNS is not updating for ONE address for some reason (it does for all others)

     

  11. 3 hours ago, binhex said:

    nothing obviously wrong that i can see, what if you try curl/wget on the host to the web ui?, if that works then your issue is blocking from/to the host from/to the internet.

     

    Thanks... I confess I have no idea hgow to curl/wget to the webui.... could you (please) spoonfeed me a bit :-) ?? I can of course SSH to the remote host. 

  12. On 1/17/2023 at 2:54 PM, fat_flying_pigs said:

    Figured out magnet links via the web ui, so posting if it might help others utilizing the forum search. They are tricky, even more so if you are using chrome (on Win 10, probably 11 too).


    Magnet requires QBT running the web ui in HTTPS. To do this we need a self signed cert.
    Pop open a console for this docker container, then make a new directory with `mkdir /config/qBittorrent/config/ssl` and enter it with `cd /config/qBittorrent/config/ssl`. Create a new self signed key with `openssl req -new -x509 -nodes -out server.crt -keyout server.key`. Enter whatever you want when it prompts; don't enter empty strings.


    Next, fix the damn folder perms! `chown -R nobody:user /config/qBittorrent/config/ssl`. Otherwise the files/folder are owned by root:root and QBT will error later on. Enter the container, Option -> Web UI, then:

     

    1. Check "Use HTTPS instead of HTTP"
    2. For Certificate, enter: /config/qBittorrent/config/ssl/server.crt
    3. For Key, enter: /config/qBittorrent/config/ssl/server.key

    Restart the container, and use the HTTPS in the url. The browser might complain about an 'unsafe' website, click advanced and proceed anyway. If you see the UI, all is good.


    Next we have to associate the magenet link. In the web ui, click Tools -> Register to handle magnet links. The browser should prompt something, accept it. If you're running firefox, that should be it - you can click a mag torrent link and it should work (there might be another browser prompt). I recommend you ctrl+click the mag link, so it opens it in another tab. If you're using chrome however... you will need to do the below.


    https://github.com/qbittorrent/qBittorrent/issues/12605#issuecomment-1385058678

     

     

    I'm assuming there is a similar work around on on mac/linux if you are using chrome. Annoying this doesn't work eaiser, and there is this much work involved. Hope this helps!

     

    Have to wonder if this could/should be built into container rather than something that has to be (with respect) bodged on?

    • Group 33 10:00.3 [RESET] 1022:149c USB controller: Advanced Micro Devices, Inc. [AMD] Matisse USB 3.0 Host Controller

     

    This one WONT pass so has to be used for unRAID USB stick

     

    • Group 20 03:08.0 [RESET] 1022:57a4 PCI bridge: Advanced Micro Devices, Inc. [AMD] Device 57a4 0a:00.0 [RESET] 1022:1485 Non-Essential Instrumentation [1300]: Advanced Micro Devices, Inc. [AMD] Starship/Matisse Reserved SPP 0a:00.1 1022:149c USB controller: Advanced Micro Devices, Inc. [AMD] Matisse USB 3.0 Host Controller 0a:00.3 [RESET] 1022:149c USB controller: Advanced Micro Devices, Inc. [AMD] Matisse USB 3.0 Host Controller

     

    I cannot pass these controllers separately (i.e. to two diff VMs) ... can anyone with AMD pass those separately? Would appreciate knowing how (did you need ACS patch?)

     

×
×
  • Create New...