TheGrownUpGeek

Members
  • Posts

    14
  • Joined

  • Last visited

Posts posted by TheGrownUpGeek

  1. Issue with HW decoding - Please Help

     

    Hi I have had the container installed for a while and added a variable to the container for a device and set the variable to /dev/dri to enable hardware decoding. This enabled the preferred Hardware decoders to be seen by emby (see attached screenshot), However when transcoding (or converting a file as part of an offline download to ios) the GPU is not being used and instead the CPU utilisation for cores allocated to the container max out at 100%.  The conversions are taking a long time to complete (about an hour for a film).  CPU is an AMD Ryzen 7 5700G with Radeon Graphics and I am trying to use the APU for transcoding.  Can you advise how i can enable HW decoding and verify its operation for media? I have also attached the CPU usage and the GPU usage graphs during a conversion below.

     

    Thanks in advance 

    Screenshot 2022-11-21 at 23.06.16.jpg

    Screenshot 2022-11-21 at 23.17.02.jpg

    Screenshot 2022-11-21 at 23.17.15.jpg

  2. Hi,

     

    I have built a Windows 11 Gaming VM, In Unraid I have Isolated the following CPUs & Threads  

     

    3/11, 4/12, 5/13, 6/14, 7/15

     

    I have allocated the paired cores and thread 3-6 and 11-14 to the VM.  After reading some performance optimisation posts I am trying to pin Core 7 & thread 15 to the emulator for the passed through GPU that i have.  Below is the VM XML extract. It has correctly allocated vcpu to the physical cores & threads.

     

    <vcpu placement='static'>8</vcpu>
      <cputune>
        <vcpupin vcpu='0' cpuset='3'/>
        <vcpupin vcpu='1' cpuset='11'/>
        <vcpupin vcpu='2' cpuset='4'/>
        <vcpupin vcpu='3' cpuset='12'/>
        <vcpupin vcpu='4' cpuset='5'/>
        <vcpupin vcpu='5' cpuset='13'/>
        <vcpupin vcpu='6' cpuset='6'/>
        <vcpupin vcpu='7' cpuset='14'/>
        <emulatorpin cpuset='7,15'/>
      </cputune>

     

    After adding the emulator pin line when the VM is running i see some activity on core 7 but no activity on thread 15.  Is that correct or is my emulator pin command wrong?  A second question is emulator pinning worth while to increase the performance of the VM by making sure emulation instructions and commands for passed through hardware operate on physical cores isolated from the VM and Unraid?  Is there anything else i can do to increase gaming performance for a windows 11 gaming VM?

     

    Thanks in advance.

  3. Hi All, having an issue with the Nginx-Proxy-Manager-Official container. I have the container running on a custom network (br0) with a static private ip set.  When launching the container i change the values for the http and https port to 8080 and 80443 respectively, however when the container runs it is still opening these ports as their default 80 & 443 values.  Has anyone seen this issue before and if so what was the fix?

    Screenshot 2022-01-26 at 22.52.02.png

    Screenshot 2022-01-26 at 22.55.22.png

  4. 56 minutes ago, Econaut said:

     

    I can share what I remember:

     

    Use official Emby docker. Put this in the config:

    I recall it just working essentially. If it's not for you, share some of your details and I can compare.

    image.png

    I get the following error when starting the container. Did you need to do anything in advance to isolate the APU or add drivers or did it just work?

     

    docker: Error response from daemon: error gathering device information while adding custom device "/dev/dri": no such file or directory.
     

  5. On 9/23/2021 at 7:57 PM, Econaut said:

    Yup that one worked for me. Also used the Emby docker (official) and used the same Device passthrough /dev/dri - select VAAPI in server config. Both worked great for me. Using a 5700G. Tried with official Plex docker and that was a failure.

    @econaut could you share in full the steps required to get this working in emby for HW transcoding on a 5700?

  6. On 1/4/2022 at 1:24 PM, Asian23 said:

    In the meantime, I'm still using the 5700G APU for plex docker which is feasible. But if we could get the 5700G passthrough working I could literally run 2 gaming VMs.......

    Hi @Asian23, this is the use case I am looking at, would you mind sharing the steps you took to pass through the APU to the plex docker for transcoding and also confirm that was on the x570?  What is the transcoding performance like compared to before in Plex?

  7. Hi I have noticed an issue when using this container in custom Bridged mode.  The container starts and the vpn connection is established however I am unable to access the deluge webui.  If I set my browser to use privoxy as a proxy I am able to access the webui fine, however when I remove the proxy config from my browser I am again unable to access the webui for deluge.  All other services continue to function as normal and a curl of ifconfig.io shows that all traffic is routed through the VPN service as expected.

     

    How can this be resolved?  Happy to provide any further info as needed. It seems crazy to have to enable the proxy in my browser just to get to the webui.

     

    Checking the supervisord it appears as below that the webui is being mapped to the internal docker network address and not 192.168.1.1 as it should be 

     

    [debug] VPN IP is 10.7.0.6
    [debug] Deluge IP is 10.7.0.6

     

  8. Hi all,

     

    I have been searching for a while and have found conflicting answers so trying to get a definitive response,  My unraid server is headless and i only connect to the admin gui remotely or via SSH so my understanding is that it is possible to pass through the integrated video controller to a VM (I do have a discreet graphics card but that is already in use in a gaming VM, i would like to pass through the integrated GPu for some video rendering and conversion tasks in a separate VM.)  I have a ryzen 3400G CPU that has built in vega 11 graphics i wondered if it would be possible to pass this through to a VM.  Having looked at the awesome @SpaceInvaderOne video on breaking up IOMMU Groups i think this may be possible but thought i would ask first as i want to avoid breaking anything in the array. I have tried the simple option of enabling PCIe ACS override: in the VM Manager Settings window however that does not split out the adapter into its own group which prevents me starting

     

    Below is the IOMMU group that contains the Vega VGA adapter

     

    IOMMU group 1:	[1022:1452] 00:08.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 00h-1fh) PCIe Dummy Host Bridge
    	[1022:15db] 00:08.1 PCI bridge: Advanced Micro Devices, Inc. [AMD] Raven/Raven2 Internal PCIe GPP Bridge 0 to Bus A
    	[1022:15dc] 00:08.2 PCI bridge: Advanced Micro Devices, Inc. [AMD] Raven/Raven2 Internal PCIe GPP Bridge 0 to Bus B
    	[1002:15d8] 06:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Picasso (rev c8)
    	[1002:15de] 06:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Raven/Raven2/Fenghuang HDMI/DP Audio Controller
    	[1022:15df] 06:00.2 Encryption controller: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 10h-1fh) Platform Security Processor
    	[1022:15e0] 06:00.3 USB controller: Advanced Micro Devices, Inc. [AMD] Raven USB 3.1
    	[1022:15e1] 06:00.4 USB controller: Advanced Micro Devices, Inc. [AMD] Raven USB 3.1
    	[1022:15e3] 06:00.6 Audio device: Advanced Micro Devices, Inc. [AMD] Family 17h (Models 10h-1fh) HD Audio Controller
    	[1022:7901] 07:00.0 SATA controller: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] (rev 61)

    I have 2 questions;

     

    1.  If i were to append the syslinux with the command pcie_acs_override=id:1022:15d8 would that break out the device so that it can be passed through to a VM and what if any are the downsides of breaking the IOMMU groups assigned by the system?

    2. Is there a different method that will allow the Integrated GPU to be used and if so how can this be achieved?

     

    Thanks in advance and if you require any further info let me know.

  9. 1 hour ago, strike said:

    No, this container does not support this at this time. You should be able to use desktop/thin client to get access though. I have mine working on windows at least (with the old v1), on custom br0 with it's own IP.. AFAIK there is no finished windows/mac client ready for v2 yet, but there are instructions on how to build it on your own, I just haven't had the time to do it yet. 

    Ok cool thanks i will do it that way, thanks for the reply

     

  10. Hi I have an issue when deploying this container.  When deployed using the bridge interface all works and the web ui is accessible.  If i use a custom bridge interface and specify a manual ip the container runs and functions as expected (privoxy works, downloads added via other sources work when in the console curl ifconfig.io returns the address of the VPN server i am connecting to).  However when using a custom bridge interface to allow the container to use its own address the web ui cannot be accessed either from within unraid or directly via http://[containerIP]:8112.  Should this container support the use of custom IP address on a container for accessing the web ui?

  11. Hi I have an issue when deploying this container.  When deployed using the bridge interface all works and the web ui is accessible.  If i use a custom bridge interface and specify a manual ip the container runs and functions as expected (privoxy works, downloads added via other sources work when in the console curl ifconfig.io returns the address of the VPN server i am connecting to).  However when using a custom bridge interface to allow the container to use its own address the web ui cannot be accessed either from within unraid or directly via http://[containerIP]:9080.  Should this container support the use of custom IP address on a container for accessing the web ui?