Skip to main content

Deploy Farmer

Prerequisites​

The drives must already be prepped using the method in this wiki:

Portainer is not required but highly recommended as it provides a nice GUI front end to Docker. Make sure to complete the last section "Keep Ubuntu VM Running"

The Node must be deployed before deploying the Farmer

Install the Latest GPU Drivers (NVIDIA) in order to use GPU Plotting

Install the NVIDIA Toolkit for Docker

Resources​

The following are relevant resources:

  1. Port Forwarding (required)
  2. Official Autonomys Wiki
  3. GPU Plotting Requirements

Deploy Autonomys Docker Networks​

For organizational purposes a network will be created for Autonomys docker containers. If your Node and Farmer will be on the same host this network should already have been created when deploying the Node. Skip to the next section if it already exists.

Right click on the Windows start menu and then select "Terminal". Then paste in and execute the following command

docker network create `
  --driver bridge `
  --subnet 172.25.0.0/16 `
  --gateway 172.25.0.1 `
  autonomys-network

This creates a new network called "autonomys-network" with the specified subnet and gateway. Close the Terminal as it is no longer needed.

Alternately, a network can be created via Portainer by selecting "Networks" > "Add Network" and then setting the following values:

  • Name: autonomys-network
  • Driver: bridge
  • Subnet: 172.25.0.0/16
  • Gateway: 172.25.0.1

Then click "Create Network"

Identify Disks and Size​

If the disk path and size are not known check them in WSL. Press WIN + x, then select "Terminal" and then open up wsl with wsl. Now display the disks with:

df --block-size=G | grep autonomys

The above command assumes you mounted the drives in a subfolder of autonomys.

Deploy Farmer​

Open up Portainer, go to "Stacks" and then click "Add Stack"

add-stack

For "Name" enter "autonomys-farmer". Navigate to the Docker farmer.yaml then copy the code and paste it into the stack file. Update the following:

  1. image: Update the image to the latest available
  2. volumes: Update the path to the Farm(s)
  3. --node-rpc-url: If the Node and Farmer are on the same host leave this. Otherwise, change this to the internal IP of the host running the Node.
  4. --reward-address: Update to your reward address
  5. path=: Update the path and size for the Farm(s) added.
  6. TZ=: Update this to your own local timezone

Once everything is updated, scroll to the bottom of the page and click "Deploy the stack".

Post Deployment​

Piece Cache​

Once the Farmer is deployed the Farms will be created and the piece cache will begin to sync. This process could take some time depending on the total disk space and size of the blockchain. It may appear like the Farmer will start plotting, but typically progress will not be made until the Farmer piece cache has synced.

GPU Plotting​

If a compatible GPU is detected, it will be used and the CPU will be disabled.

Replotting​

Once the initial plotting is finished, the Farmer will likely begin to replot as sectors become stale over time. The replotting should slow as the network becomes larger.

Node + Farmer Stack​

If the Node and Farmer will be on the same host, it can be easier to run them in the same stack file. To view an example stack file using this method, navigate to the Docker node-farmer-combined.yaml

The Node and Farmer services are almost identical, with the exception that a depends_on section has been added to ensure the Node has fully started before the Farmer starts. This is because the Node must be running before the Farmer.