io.net | Worker Node Setup Guide

io.net is building an enterprise-grade decentralized computing network that allows machine learning engineers to access distributed cloud clusters at a small fraction of the cost of comparable centralized services.

With io.net, teams can scale their workloads across a network of GPUs with minimal adjustments. The system handles orchestration, scheduling, fault tolerance, and scaling and supports a variety of tasks such as preprocessing, distributed training, hyperparameter tuning, reinforcement learning, and model serving. It is designed to serve general-purpose computation for Python workloads, with an emphasis on serving AI/ML workloads.

Site | Docs | Roadmap | X | Discord | LinkedIn

Before interact with any project, please, DYOR about the project:
1) DYOR guide
2) DYOR practices

Bonus: Reward program

Server Requirements

For this node You’ll require high quality GPU or CPU.
You can check the list of supported devices here.

It’s better to use dedicated server with required CPU or GPU, but You can try Your luck some VPS.

Check possible servers here:
- Hetzner
- Cityhost

For a test purpose I have rented CPX41 server on Hetzner. But the chances of getting rewards on this VPS is low.

Prerequisite

Installation for this worker node made easy.
Will only need Docker and docker-compose, but there is a script for automatic installation of the node and docker/docker compose, will just prepare our server for this installation.

1) Update packages

sudo apt-get update && apt-get upgrade -y

2) Install curl

sudo apt install curl

Installation

1) Go to cloud.io.net

2) Connect with Your Google/X account or e-mail.

3) Move cursor to the left upper corner and click the “Worker” button.

4) Connect new worker.

Now You’ll need to follow all the steps provided by the team.

5) Select Operating System “OS”. Choose Linux.

6) Choose the supplier you wish your hardware to be grouped on.

7) Add a unique name for your device.

The ideal format would be similar to the following: My-Test-Device

8) Choose device type.
If you choose GPU Worker and your device doesn’t have a GPU, the setup will fail.

For this test I got CPU.

9) Create io-net folder:

mkdir io-net

10) Move to io-net folder:

cd io-net

11) Download the setup script:

curl -L https://github.com/ionet-official/io-net-official-setup-script/raw/main/ionet-setup.sh -o ionet-setup.sh

12) Make ionet-setup.sh file executable:

chmod +x ionet-setup.sh

13) Run ionet-setup.sh script:

./ionet-setup.sh

14) Download binary:

curl -L https://github.com/ionet-official/io_launch_binaries/raw/main/launch_binary_linux -o launch_binary_linux

15) Make binary executable:

chmod +x launch_binary_linux

16) Run the command to connect device. Just copy/paste code provided by the team.

Get crptcpchk’s stories in your inbox

Join Medium for free to get updates from this writer.

Remember me for faster sign in

Check that device name is the same as You entered in this step 7.

This is how successful installation looks like:

17) Check that both containers got installed:

docker ps

Should look like this:

If there’re no containers or only one got installed, follow steps 14–16 once again.

Now wait till You device is connected. Should look like this:

18) Follow provided link to the device page.
You should see a dashboard of You device, where You can track performance of the node.

Performance monitoring

Go to this page.

It should show Your connected devises with their performance as showed above.

P.S. If your devices didn’t show up on this page, You need to enter device name and the link will pop up. Click on the device_id that showed up to open device dashboard.

Delete Node

1. Run:

docker ps

2) Run for both containers:

docker stop <CONTAINER ID>

*CONTAINER ID shoud be change to the IDs of Your containers.

3) Move to the root folder:

cd

4) Delete io-net folder:

rm -rf io-net

Links

Follow for more guides on nodes, drops and other useful stuff.

psi.crypto Discord: https://discord.gg/9kvtEQwtqJ

X: https://twitter.com/the01100011

Telegram: https://t.me/psicrypto

Farcaster: @crptcpchk

HackenDAO: https://discord.gg/hacken

RawBox: https://twitter.com/rawbox_tech