Skip to main content

Join as Compute Provider

Bacalhau is a peer-to-peer network of compute providers that will run jobs submitted by users. A Compute Provider (CP) is anyone who is running a Bacalhau compute node participating in the Bacalhau compute network, regardless of whether they are hosting any Filecoin data.

This section will show you how to configure and run a bacalhau node and start accepting and running jobs.

To bootstrap your node and join the network as a CP you can leap right into the Ubuntu 22.04 quick start below, or find for more setup details in these guides:

info

If you run on a different system than Ubuntu, drop us a message on Slack! We'll add instructions for your favorite OS.

Quick start (Ubuntu 22.04)

Estimated time for completion: 10 min.

Tested on: Ubuntu 22.04 LTS (x86/64) running on a GCP e2-standard-4 (4 vCPU, 16 GB memory) instance with 40 GB disk size.

Prerequisites

  • Docker Engine - to take on Docker workloads
  • Connection to IPFS - for storing job's results
  • Firewall - to ensure your node can communicate with the rest of the network

Install docker

To run docker based workloads, you should have docker installed and running.

If you already have it installed and want to configure the connection to Docker with the following environment variables:

  • DOCKER_HOST to set the url to the docker server.
  • DOCKER_API_VERSION to set the version of the API to reach, leave empty for latest.
  • DOCKER_CERT_PATH to load the TLS certificates from.
  • DOCKER_TLS_VERIFY to enable or disable TLS verification, off by default.

If you do not have Docker on your system, you can follow the official docker installation instructions or just use the snippet below:

# install dependencies
sudo apt-get update
sudo apt-get install \
ca-certificates \
curl \
gnupg \
lsb-release

# add package repo and key
sudo mkdir -p /etc/apt/keyrings
curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo gpg --dearmor -o /etc/apt/keyrings/docker.gpg
echo "deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.gpg] https://download.docker.com/linux/ubuntu \
$(lsb_release -cs) stable" | sudo tee /etc/apt/sources.list.d/docker.list > /dev/null

# install docker
sudo apt-get update
sudo apt-get install docker-ce docker-ce-cli containerd.io docker-compose-plugin

Now make Docker manageable by a non-root user:

sudo groupadd docker
sudo usermod -aG docker $USER

Ensure IPFS is running

We will need to connect our bacalhau node to an IPFS server so we can run jobs that consume CIDs as inputs.

You can either install IPFS and run it locally or you can connect to a remote IPFS server.

In both cases - we should have an IPFS multiaddress for our IPFS server that should look something like this:

export IPFS_CONNECT=/ip4/10.1.10.10/tcp/80/p2p/QmVcSqVEsvm5RR9mBLjwpb2XjFVn5bPdPL69mL8PH45pPC
caution

The multiaddress above is just an example - you need to get the multiaddress of the IPFS server you want to connect to.

To install a single IPFS node locally on Ubuntu you can follow the official instructions, or follow the steps below.

We advise to run the same IPFS version as the Bacalhau main network. Pick that with the command below:

export IPFS_VERSION=$(wget -q -O - https://raw.githubusercontent.com/filecoin-project/bacalhau/main/ops/terraform/production.tfvars | grep --color=never ipfs_version | awk -F'"' '{print $2}')

Install:

wget "https://dist.ipfs.tech/go-ipfs/${IPFS_VERSION}/go-ipfs_${IPFS_VERSION}_linux-amd64.tar.gz"
tar -xvzf "go-ipfs_${IPFS_VERSION}_linux-amd64.tar.gz"
cd go-ipfs
sudo bash install.sh
ipfs --version

Configure:

sudo mkdir -p /data/ipfs
export IPFS_PATH=/data/ipfs
sudo chown $(id -un):$(id -gn) ${IPFS_PATH} # change ownership of ipfs directory
ipfs init

Now launch the IPFS daemon in a separate terminal (make sure to export the IPFS_PATH environment variable there as well):

ipfs daemon
info

If you want to run the IPFS daemon as a systemd service, here's an example systemd service file.

Don't forget we need to fetch an IPFS multiaddress pointing to our local node. Use the following command to print out a number of addresses.

ipfs id
{
"ID": "12D3KooWNhRU6H1eeqvT1jQAXFAdFDvT5H4AEGmHV7N1cYbzYh1F",
"PublicKey": "CAESIL9gmDyR6IgM7ym1JmJ8JKL7NvEgIEGaWwssanl1ieuW",
"Addresses": [
"/ip4/10.128.0.11/tcp/4001/p2p/12D3KooWNhRU6H1eeqvT1jQAXFAdFDvT5H4AEGmHV7N1cYbzYh1F",
"/ip4/10.128.0.11/udp/4001/quic/p2p/12D3KooWNhRU6H1eeqvT1jQAXFAdFDvT5H4AEGmHV7N1cYbzYh1F",
"/ip4/127.0.0.1/tcp/4001/p2p/12D3KooWNhRU6H1eeqvT1jQAXFAdFDvT5H4AEGmHV7N1cYbzYh1F",
"/ip4/127.0.0.1/udp/4001/quic/p2p/12D3KooWNhRU6H1eeqvT1jQAXFAdFDvT5H4AEGmHV7N1cYbzYh1F",
...
],
"AgentVersion": "go-ipfs/0.12.2/",
"ProtocolVersion": "ipfs/0.1.0",
"Protocols": [
"/ipfs/bitswap",
...
"/p2p/id/delta/1.0.0",
"/x/"
]
}

I pick the record that combines 127.0.0.1 and tcp but I replace port 4001 with 5001:

export IPFS_CONNECT=/ip4/127.0.0.1/tcp/5001/p2p/12D3KooWNhRU6H1eeqvT1jQAXFAdFDvT5H4AEGmHV7N1cYbzYh1F

Configure firewall

To ensure that our node can communicate with other nodes on the network - we need to make sure the 1235 port is open.

(Optional) To ensure the cli can communicate with our node directly (i.e. bacalhau --api-host <MY_NODE_PUBLIC_IP> version) - we need to make sure the 1234 port is open.

Firewall configuration is very specific to your network and we can't provide generic instructions for this step but if you need any help feel free to reach out on Slack!

Install the bacalhau binary

Install the bacalhau binary to run bacalhau serve.

info

If you want to run Bacalhau as a systemd service, here's an example systemd service file.

Run bacalhau

Now we can run our bacalhau node:

LOG_LEVEL=debug bacalhau serve \
--ipfs-connect $IPFS_CONNECT

With the command above our node joins the Bacalhau network, congrats! 🎉

Check your node works

Even though the cli (by default) submits jobs to the bootstrap nodes, each node listens for events on the global network and possibly bids for taking a job: your logs should therefore show activity of your node bidding for incoming jobs.

To quick check your node runs properly, let's submit the following dummy job:

bacalhau docker run ubuntu echo Test

If you see logs of your computenode bidding for the job above it means you've successfully joined Bacalhau as a Compute Provider!

What's next?

At this point you probably have a number of questions for us. What incentive should you expect for running a public Bacalhau node? Please contact us on Slack to further discuss this topic and for sharing your valuable feedback.