Skip to content

Latest commit

 

History

History
287 lines (201 loc) · 7.7 KB

clients.md

File metadata and controls

287 lines (201 loc) · 7.7 KB
Raw

Clients

The Cortex repository is a fork of Go Ethereum which contains protocol changes to support machine learning model inference. This implements the Cortex network, which maintains a separate ledger from the Ethereum network, for several reasons, the most immediate of which are AI inference support and the deterministic inference consensus.

Requirements

To ensure that your Cortex client runs gracefully, please check your system meets the following requirements:

  • System: Linux Ubuntu 16.04+
  • CPU: Xeon processor 16 Cores
  • GPU: Nvidia GPU 1060
  • RAM: 32GB
  • Space: 2TB (the size of the blockchain increases over time)
  • CUDA version: 9.2+
  • CUDA driver: 396+
  • Compiler: Go 1.10+, GCC 5.4+

note: Currently, we only officially support Linux (Ubuntu), but we are actively investigating development for other operating systems and platforms(e.g., macOS, Ubuntu, Windows, Fedora).

Building Cortex clients requires a Go (version 1.10 or later), a C compiler, and a CUDA (version 9.2 or later). We will guide you through the Installation section.

Installation

Before we begin installing Cortex, we need to get some dependencies for your system.

Install NVIDIA Driver

Remove old driver

sudo apt-get remove --purge nvidia*
sudo apt-get install build-essential freeglut3-dev libx11-dev libxmu-dev libxi-dev libgl1-mesa-glx libglu1-mesa libglu1-mesa-dev
sudo nano /etc/modprobe.d/blacklist-nouveau.conf

Paste the following content into the text editor and save blacklist-nouveau.conf,

blacklist nouveau
options nouveau modeset=0

Download the driver - Option 1 - via Nvidia.com

wget http://us.download.nvidia.com/XFree86/Linux-x86_64/410.93/NVIDIA-Linux-x86_64-410.93.run
chmod +x NVIDIA-Linux-x86_64-410.93.run

Download the driver - Option 2 - via PPA repository

sudo add-apt-repository ppa:graphics-drivers/ppa
sudo apt-get update

Press Ctrl + Alt + F1 to enter tty1 console,

sudo update-initramfs -u
sudo service lightdm stop
sudo ./NVIDIA-Linux-x86_64-410.93.run –no-opengl-files
(do not apply xorg config here!!!!)

Then

sudo service lightdm start

Press Ctrl + Alt + F7 to go back tty7 interface.

Install CUDA (without NVIDIA Driver)

Option 1 - Install CUDA 10.0 (Without NVIDIA Driver)

wget https://developer.nvidia.com/compute/cuda/10.0/Prod/ local_installers/cuda_10.0.130_410.48_linux
mv cuda_10.0.130_410.48_linux cuda_10.0.130_410.48_linux.run
sudo sh cuda_10.0.130_410.48_linux.run
sudo ldconfig /usr/local/cuda/lib64
(IMPORTANT: don't install driver here!!!)

// Add two lines to ~/.bashrc
echo 'export PATH=/usr/local/cuda-10.0/bin${PATH:+:${PATH}}' >> ~/.bashrc
echo 'export LD_LIBRARY_PATH=/usr/local/cuda-10.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}' >> ~/.bashrc
source ~/.bashrc

Option 2 - Install CUDA 9.2 (Without NVIDIA Driver)

wget https://developer.nvidia.com/compute/cuda/9.2/Prod2/local_installers/cuda_9.2.148_396.37_linux
mv cuda_9.2.148_396.37_linux cuda_9.2.148_396.37_linux.run
sudo sh cuda_9.2.148_396.37_linux.run
sudo ldconfig /usr/local/cuda/lib64
(IMPORTANT: don't install driver here!!!)

// Add two lines to ~/.bashrc
echo 'export PATH=/usr/local/cuda-0.2/bin${PATH:+:${PATH}}' >> ~/.bashrc
echo 'export LD_LIBRARY_PATH=/usr/local/cuda-9.2/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}' >> ~/.bashrc
source ~/.bashrc

Install Go

Optnion 1 - Install via Google.com

wget https://dl.google.com/go/go1.11.5.linux-amd64.tar.gz
sudo tar -C /usr/local -xzf go1.11.5.linux-amd64.tar.gz
echo 'export PATH="$PATH:/usr/local/go/bin"' >> ~/.bashrc
source ~/.bashrc

Optnion 2 - Install via package manager

sudo apt install golang-1.11

Make sure you have installed correctly

go version

Change Linux Open File Limit

you will have to edit the following file:

vi /etc/sysctl.conf

Add the following line:

fs.file-max=500000

Of course, you can change the number per your needs. To verify the changes again use:

cat /proc/sys/fs/file-max

Users will need to logout and login again for the changes to take effect. If you want to apply the limit immediately, you can use the following command:

sysctl -p

Set User Level Open File limits in Linux

For that purpose, as user root, you will need to edit the following file:

vi /etc/security/limits.conf

The lines that you should add take the following parameters:

<domain>        <type>  <item>  <value>

Here is an example of setting a soft and hard limits for user cortex:

## Example hard limit for max opened files
cortex        hard nofile 409600
## Example soft limit for max opened files
cortex        soft nofile 102400

Another example of setting a soft and hard limits for all user:

## Example hard limit for max opened files
cortex        hard nofile 409600
## Example soft limit for max opened files
cortex. /      soft nofile 102400

Building the source

Clone the source (Need permission)

git clone https://github.com/CortexFoundation/CortexTheseus.git --branch dev
(with git accessable key)

cd CortexTheseus

Once the dependencies are installed, run

make -j cortex

Save the executable file

sudo mkdir -p /serving/cortex-core/bin
sudo chmod 777 /serving/cortex-core/bin
cp build/bin/cortex /serving/cortex-core/bin/cortex

The compiled binary files are located in the ./build/bin

./build/bin/cortex

Run the fullnode

Fullnode executables directly

./cortex --port 37566 --rpc --rpccorsdomain '*' --rpcport 30089 --rpcaddr 127.0.0.1 --rpcapi web3,personal,ctxc,miner,net,txpool --verbosity 4 --gcmode archive

Create bash script

rm /serving/cortex-core/bin/cortex.sh
sudo nano /serving/cortex-core/bin/cortex.sh

Create /serving/cortex-core/bin/cortex.sh

#!/bin/bash
export PATH=/usr/local/cuda-10.0/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
/serving/cortex-core/bin/cortex --port 37566 --rpc --rpccorsdomain '*' --rpcport 30089 --rpcaddr 127.0.0.1 --rpcapi web3,ctxc,miner,net,txpool,personal --verbosity 4 --gcmode archive

Make the script executable

sudo chmod +x /serving/cortex-core/bin/cortex.sh

config

[program:cortexnode]
directory=/serving/cortex-core/bin/
command=bash /serving/cortex-core/bin/cortex.sh
autostart=true
autorestart=true
startsecs=5
stderr_logfile=/tmp/cortex_fullnode_stderr.log
stdout_logfile=/tmp/cortex_fullnode_stdout.log

check running status

sudo supervisorctl tail cortexnode stdout

Executables

Command

use the parameter directly.

parameter description
help Print Cortex binary help statements.
init Init Cortex node from a genesis configuration file

Options

use "--" prefix with the parameter.

parameter description
storage.dir Set the directory of Cortex storage layer. By default, $HOME/.cortex/storage the directory is the storage dir.
datadir Set Cortex binary data directory. By default, $HOME/.cortex/ is data directory.
port Set Cortex binary listening port.
bootnodes [Optional] Set bootnodes of Cortex blockchain.
verbosity Set logging level to print, by default is 3, range in [1, 5], which represent Error, Warn, Info, Debug, Trace.