[GUIDE] Installing ARM in a Proxmox Container

[SylvainMT]

Installation steps of Automatic Ripping Machine (Docker Install) inside a Proxmox LXC container

This guide will install the Automatic Ripping Machine (ARM) inside a Proxmox Container LXC using the docker install script.

Storage Considerations

In this setup I use a shared volume for ARM to deposit completed files into which is shared with other containers installed on for use, in my case Jellyfin. LXD containers are persistent which means that settings can be saved in the container's volume. In my setup Proxmox is running on an SSD which is the fastest storage available so my transcode and raw folders are remaining inside the container. That does mean however that I needed to give enough storage space to my container to work. (100 GiB).

Tested Environment

• Proxmox VE 8.0.4
• Computer
  • Asus P5QL-E Motherboard
  • Core(TM) 2 Quad CPU Q6600 @ 2.40GHz
  • 16 GiB of DDR2 Ram
  • Pioneer DVD-RW DVR-107D Optical Drive
  • NVidia GeForce GTX 1050Ti (Asus Strix) (Note that this guide doesn't include GPU Passthrough... yet...)

Procedure

First one needs to prepare the Proxmox Host and get Information

1. Log into Proxmox as root (makes the rest easier) and open your node's Shell ("Datacenter" -> [YourNodeName] -> ">_ Shell" This should give you a Shell screen logged in as root.
2. apt update && apt upgrade -y
3. Install lsscsi apt install lsscsi
4. lsscsi -g
   The command returns a screen like below, write down your optical drive's SR id (in the blue box of the image) and SG id (in the green box of the image)
   
5. ls -l /dev/sr* && ls -l /dev/sg*
   This command will list your drive's major (in the blue box) and minor (in the green box) number for both the SR id, the SG id as well as the device's category (the letter "b" and "c" in the yellow box) Record all three values.
   
6. Download the Debian-12 Container image
   1. Choose "local" storage
   2. "CT Templates"
   3. "Templates"
   4. Search for "Debian"
   5. Select "debian-12-standard"
   6. Download
7. Create the container
   1. "Create CT"
   2. Enter a CTID, Hostname and Password of your choice.
   3. Deselect "Unprivileged container" IMPORTANT
   4. "Next"
   5. Choose the previously downloaded Debian-12 template and click "Next"
   6. Give plenty of room for your root storage I choose 100 GiB. It will become your scratch drive (We will mount a shared volume later)
   7. Decide how many cores to let your container access. I suggest as many as you can, I gave it all four. Minimum of 2 is recommended.
   8. Give your container access to some RAM, I went with 2GiB (2048MiB) and 512MiB Swap Drive
   9. Enter your network details. I suggest a static IP as this is the IP you will use to access ARM
   10. Enter your DNS settings (or just use the Host Settings)
   11. Confirm everything but do not start the container yet
   12. Once your container is done building, click on the container in the UI > Options > Features > Edit. Add a checkbox next to Nesting. Click OK.
8. Passing through the Optical drive(s) to your container.
   1. Go back to your node's Shell interface
   2. Edit the container's .conf file nano /etc/pve/lxc/<Your Container ID>.conf (It should look similar to below)

arch: amd64
cores: 4
hostname: arm
memory: 2048
net0: name=eth0,bridge=vmbr0,firewall=1,gw=<Your Gateway IP>,hwaddr=<Your MAC Address>,ip=<Your IP>,type=veth
ostype: debian
rootfs: local-lvm:vm-300-disk-0,size=100G
swap: 512

~
3. Add the following lines.

lxc.apparmor.profile: unconfined
lxc.cap.drop:
lxc.autodev: 1
lxc.mount.auto: sys:rw

~
4. Add the following lines for each of the Optical Drives you are passing to the Container (Replace the values inside <Value> to what is appropriate for your situation)

lxc.cgroup2.devices.allow: <SR Device Category> <SR Major Number>:<SR Minor Number> rwm
lxc.mount.entry: /dev/sr<SR ID Number> dev/sr<SR ID Number> none bind,create=file,optional 0 0
lxc.cgroup2.devices.allow: <SG Device Category> <SG Major Number>:<SG Minor Number> rwm
lxc.mount.entry: /dev/sg<SG ID Number> dev/sg<SG ID Number> none bind,create=file,optional 0 0

This is my configuration

lxc.cgroup2.devices.allow: b 11:0 rwm
lxc.mount.entry: /dev/sr0 dev/sr0 none bind,create=file,optional 0 0
lxc.cgroup2.devices.allow: c 21:0 rwm
lxc.mount.entry: /dev/sg0 dev/sg0 none bind,create=file,optional 0 0

7. Start your container and log-in to your container's Console ("Container Name" > ">_ Console") as root
   1. `apt update && apt upgrade -y'
   2. apt install lsscsi
   3. Confirm that your Optical Drive is passed through with lsscsi -g confirm that the output for your Optical Drives are the same as they are on the Proxmox Host. If not check your LXC Config file and restart the container.
   4. Confirm that you have access to the drive with ls -l /dev/sr* && ls -l /dev/sg* The lines about your Optical Drives should be identical as they were on the Proxmox Host. If not check your LXC Config file and restart the container.
   5. Create the Arm User and Group. IMPORANT To be able to use Shared Storage Volumes, the group ID and user ID must be the same for all the users and groups (in all containers) that will used the shared volume. (So your jellyfin user and group in the jellyfin container should have the same UserID and GroupID)
      1. groupadd -g <GroupID> arm
      2. useradd -u <UserID> -g arm arm
      3. passwd arm
      4. Set a password (you will need it later)
   6. Install Docker
      1. apt install curl -y
      2. curl -sSL https://get.docker.com/ | sh
      3. Optionally Install Portainer a UI took for managing Docker containers. This step is useful if you need to do some troubleshooting
         1. docker run --restart always -d -p 9000:9000 -v /var/run/docker.sock:/var/run/docker.sock -v portainer_data:/data portainer/portainer-ce
         2. Navigate to: http://[LXC_Container_ID]:9000 to complete portainer setup.
8. Install ARM as per the wiki (But do not Start AMR Yet!, Wait to complete the "Post Install" step of the wiki guide until later.) docker
9. If the volume you intend to save the completed rips in doesn't already exist follow below, else skip to step 10.
   1. From the Proxmox UI click on your container
   2. Click on "Resources"
   3. "Add" > "Mount Point"
   4. Choose a size, set a mount point id if desired.
   5. Set the path (this is the path inside the container) to /mnt/media
   6. If you wish to include this volume in backup ensure the "backup" checkbox is selected.
   7. Click "Create"
10. In this step we make the desired volume shareable and add it to each LXC container that needs access to it.
    1. Enter your Proxmox Node Shell (and log-in as root if not already logged in)
    2. nano /etc/pve/lxc/<ContainerID>.conf here container ID is for the container that has the volume you wish to share, if you completed step 9. this is your ARM container.
    3. Find the mount point entry that equates to the container you wish to share for example mp0: local-lvm:vm-300-disk-1,mp=/mnt/media,backup=1,size=16T Hint: the "mp0" should match the Mount Point ID seen in the UI.
    4. Make it shareable by adding shared=1 to it. Copy the entire line.
    5. Edit the config file for each container you wish to share this volume with and paste the line in it. IMPORTANT The Mount Point ID 'mp#' needs to be unique within each container!!! Modify the Mount Point ID as needed! IMPORTANT To make linux permissions easier. Be sure to give each user and group, in each container the SAME <UserID> and <GroupID> how to do this is outside the scope of this guide.
11. Complete the Wiki "Post Install" steps (But do not start ARM yet!)
12. Open your Container's Console and log in as the "arm" user and add the Optical Drive and your mount point to Docker.
    1. nano start_arm_container.sh
    2. Add the optical drive (repeat for each optical drive you have)

    --device="/dev/sr<SR ID>:/dev/sr<SR ID>" \
    --device="/dev/sg<SG ID>:/dev/sg<SG ID>" \

~
3. Add your mount points these take the form of -v "<LXC Container's Path to location>:<Docker Path to location>" \ Keep in mind that here the location are from the LXC Container's and Docker's perspective respectively. (you may have to delete some lines already there or modify them...)

    -v "/home/arm:/home/arm" \
    -v "/mnt/media/Music:/home/arm/Music" \
    -v "/home/arm/logs:/home/arm/logs" \
    -v "/home/arm/media:/home/arm/media" \
    -v "/mnt/media/arm_completed:/home/arm/media/completed" \
    -v "/etc/arm/config:/etc/arm/config" \

14. Start ARM
    1. Log out from the arm account with exit
    2. Log in as root
    3. /home/arm/start_arm_container.sh

Assuming all was done correctly (and my guide is correct) you should now be able to use ARM on your Proxmox machine. Just give ARM a few seconds to start, go to the address http://<your arm ip address>:8080 log in with the default user (admin) and password (password) and once you see the arm interface pop-in a disk in your optical drive and see if it runs!

Edits

• Edited for legibility
• Noticed I forgot the steps to install docker. Added those steps to the guide.

[TallPatrick]

Suggestion: add a point in step 7 to change the console mode to shell, otherwise you can get just a black screen for your container when it first starts

[SylvainMT]

I actually didn't know that. When I do this I almost exclusively SSH in, either with Windows PowerShell or the Linux terminal, depending which computer I am working with at the time. I did intend to try and keep this guide as simple as I could, even if this is incredibly complex.

[TallPatrick]

yeah, I followed the guide solely through the in-browser console window/shell as I found that to be the simplest way to do it.

[TallPatrick]

A Step that I feel is missing is how to log in to the arm user created in step 7 to then run the docker setup script

[SylvainMT]

This step is not needed. You can install docker and run the arm setup scripts as root.

[TallPatrick]

When I did that, the script threw errors because it was trying to run sudo commands as root. This might be solved by the next comments about installing sudo as a command.

[SylvainMT]

Right I forgot that part, I got around it by creating a user for myself, installing sudo and giving it sudo privileges. Thanks for reminding me I will amend the Guide to fix this issue!

[TallPatrick]

After installing docker, a few steps are missing:

1. adding sudo as a command: apt-get install sudo -y
2. Giving the arm user the ability to use sudo: usermod -aG sudo arm
3. Adding the arm user to the docker group to be able to run docker: usermod -aG docker arm

[SylvainMT]

I see your point, but I only ever logged in as the arm user in my guide to edit the start_arm_container.sh file in step 12 and in retrospect to complete the install you don't actually need to ever log in as the arm user. I created the user account separately so that we can log in to set the settings file. (In my arm installation I actually created a handbrake_profile and mkv_profile folder to have custom settings for my ripped files (I wanted both French and English Subtitles and Audio, as well as up-scaling of my DVDs to 720p which required custom profiles and those live in the /home/arm/ directory.

[TallPatrick]

Either way, the debian image you selected doesn't seem to have sudo by default, so at least that first step I suggested seems to be needed to run the arm_setup.sh script, since it uses sudo in it.

[TallPatrick]

It also looks like the home directory for the arm user needs to manually be set up and have permissions given to it so it can write files to that location

[SylvainMT]

I'll look into using the adduser command instead of the useradd command to create the user that might fix this particular issue.

[TallPatrick]

For the note at the end of step 9, I think you mean to say to skip to step 11?

[SylvainMT]

So steps 9 and 10 deal entirely with using a shared volume.

Step 9 if you do not already have a volume you intend to share between multiple LXC containers.
Step 10 when you do have a volume you intend to share between multiple LXC containers.

If you do not intend to use shared volumes at all, then yes skip both 9 and 10 straight to 11.

[SylvainMT]

On second thought. Maybe I should extract those steps entirely and place them into their own guide. I might be trying to do too many things in this one guide here...

Opinions?

[TallPatrick]

I think either is fine. I think it's important to have them where they are, since doing those steps before running the docker container is important. I like the revised phrasing you have in your first comment here; I think it is more clear and less confusing.

[jereskuta]

Hi! Thank you for this guide, I got the ripping machine working thanks to you. Do you know how I could get the intel quicksync to work in the container?

I've passed through both /dev/dri and /dev/dri/renderD128 and I added arm to a render group on the LXC and mapped the gid to the host render group. Quicksync is still not working though. Any ideas?

[chdrsto]

Hi, any success to get quicksync up and running?
I'm having the same situation.

[SylvainMT]

@jereskuta && @chdrsto

Yes, I did get Quicksync working. It's a troublesome process since the Arm Docker release doesn't have the quick sync drivers built in. I had completely forgotten to update this.

Give me a week or so and I'll prep a guide for you guys. (Long story short, you need to go into the Docker container, and install a whole bunch of stuff, to get the QuickSync drivers (for HandBrake) in there and so some very minor changes to the ContainerID.lxc file in Proxmox and the Start_arm Docker file). This guide in the wiki: https://github.com/automatic-ripping-machine/automatic-ripping-machine/wiki/intel-qsv is where all the work inside the docker container comes from.