standalone-deploy

FATE Standalone Deployment Guide

中文

1. Description

Server Configuration.

Number	1
Configuration	8 core / 16G memory / 500G hard disk
OS	Version: CentOS Linux release 7
User	User: app owner:apps

The standalone version provides 3 deployment methods, which can be selected according to the actual situation.

• Install FATE using Docker image

• Install FATE on the host (using the compiled installer)

• Install FATE in the host (based on the source code compiled by the package)

2. Install FATE using a Docker image (recommended)

It is recommended to use a docker image, which greatly reduces the possibility of encountering problems

Note that the ${version} in the following example, please replace it with the actual version number, refer to fate.env file for the FATE version!

2.1 Pre-deployment environment check

• The host needs to be able to access the external network to pull installation packages and docker images from the public network.
• Dependent on docker, the recommended version of docker is 18.09. You can verify the docker environment with the following command: docker --version,docker start-stop and other operations please refer to docker --help
• Before executing, please check if 8080 is already occupied. If you want to execute it again, please use the docker command to delete the previous containers and images

Set the environment variables required for deployment (note that the environment variables set in the following way are only valid for the current terminal session, if you open a new terminal session, such as a new login or a new window, please set them again)

export version={FATE version for this deployment}

example:

export version=1.7.0

2.2 Pulling mirrors

2.2.1 Via the public mirror service

docker pull federatedai/standalone_fate:${version}

2.2.2 Via mirror packages

wget https://webank-ai-1251170195.cos.ap-guangzhou.myqcloud.com/fate/${version}/release/standalone_fate_docker_image_${version}_release.tar;
docker load < standalone_fate_docker_image_${version}_release.tar;
docker images | grep federatedai/standalone_fate

If you can see the image corresponding to ${version}, the image is downloaded successfully

2.3 Boot

docker run -d --name standalone_fate -p 8080:8080 federatedai/standalone_fate:${version};
docker ps -a | grep standalone_fate

If you can see that the container corresponding to ${version} is running, it starts successfully

2.4 Testing

• Enter the container

docker exec -it $(docker ps -aqf "name=standalone_fate") bash

• test item

3. Install FATE in the host (using the compiled installer)

Note that in the following example ${version}, please replace it with the actual version number, refer to fate.env file for the FATE version!

3.1 Pre-deployment environment check

Whether local ports 8080, 9360, 9380 are occupied

netstat -apln|grep 8080;
netstat -apln|grep 9360;
netstat -apln|grep 9380

3.2 Get the installation package

Download the installation package and unpack it

wget https://webank-ai-1251170195.cos.ap-guangzhou.myqcloud.com/fate/${version}/release/standalone_fate_install_${version}_release.tar.gz;
tar -xzvf standalone_fate_install_${version}_release.tar.gz

3.3 Installation

Go to the unpacked directory and use init.sh to install

The script will complete automatically:

• Install the necessary OS dependencies

• Install python36 environment

• Install pypi dependencies

• Install the jdk environment

• Configure the FATE environment variable script

• Configure fateflow

• Configure fateboard

• Install the fateboard client

  cd standalone_fate_install_${version}_release;
  sh init.sh init

3.4 Start

sh init.sh status;
sh init.sh start

3.5 Testing

• Load environment variables

source bin/init_env.sh

• test items

4. test items

4.1 Toy test

flow test toy -gid 10000 -hid 10000

If successful, the screen displays a statement similar to the following:

success to calculate secure_sum, it is 2000.0

4.2 Unit tests

fate_test unittest federatedml --yes

If successful, the screen displays a statement like the following:

there are 0 failed test

Some use case algorithms are in examples folder, please try using them.

You can also experience the algorithm process kanban through your browser by visiting: Http://${ip}:8080, ip is 127.0.0.1 or the actual ip of the local machine

5. install FATE in the host (based on the source code to compile their own package)

Please refer to standalone fate source code deployment