Installation Guide

Here are step-by-step instructions to install and use ID Fort.

1. Configuring Docker Engine

The execution of ID Fort necessitates the utilization of Docker. The installation process of Docker varies across different operating systems. To obtain guidance on installation, kindly consult the official Docker website for specific instructions. Each OS requires a configuration laid out by Docker that can be viewed by following the URLs below.

2. Nvidia Container Toolkit

Often, the consumer can utilize ID Fort over their GPU Server to boost the service performance. To cater this, Nvidia Contain Toolkit needs to be configured on the host system. Please refer to Nvidia Container Toolkit Installation Guide

Once Installed, the user needs to run the following command to ensure GPU can be utilized inside the Docker container.

Please refer to Running a Sample Workload

$ sudo docker run --rm --runtime=nvidia --gpus all ubuntu nvidia-smi

3. ID Fort Image

Each consumer is given a download link to ID Fort Docker Image. You should have been given a download link to ID Fort images. Here is the command to download the image onto your server.

📘

Appropriate Version

Make sure to download the correct version depending if you are running it on CPU or GPU.

a) Downloading Docker Image

$ curl -O <https://example.com/idfortcpu.tar>

b) Loading Docker Image

The next step for the user is to load the .tar file into Docker image.

For GPU：

$ docker load --input gpu-idfort-20240603.tar

For CPU：

$ docker load --input cpu-idfort-20240531.tar

c) Verify Docker Image

The loaded Docker image needs to be checked. This is done by listing all uploaded images over Docker.

$ docker images

The above command should send out the below output over the Docker console.

REPOSITORY   TAG        IMAGE ID       CREATED       SIZE
idfort/gpu   20240603   3f0634a5cad6   2 days ago    31.4GB
idfort       20240531   4a68eb890d54   5 days ago    18.7GB

d) Create Container

Consumer should now copy the image ID from the above response and create a Docker Container accordingly. Container can either be a CPU based Docker Container or GPU based Docker Container.

CPU Container

MEM_LIMITE：You want to limit the memory size that the container can use.

HTTP_PORT、HTTPS_PORT：The idfort container uses ports 80 and 443 internally; these are the port settings for internal to external exposure.

Here is the sh script：

MEM_LIMITE=20gb
HTTP_PORT=80
HTTPS_PORT=443
docker create --restart always -p $HTTPS_PORT:80 -p $HTTP_PORT:443 --ulimit memlock=819200000:819200000 -m $MEM_LIMITE --memory-swap $MEM_LIMITE -w /opt/idfort --entrypoint "/opt/idfort/bin/idfort" --name idfort idfort:20240531 -container 1 

GPU Container

MEM_LIMITE：You want to limit the memory size that the container can use.

HTTP_PORT、HTTPS_PORT：The idfort container uses ports 80 and 443 internally; these are the port settings for internal to external exposure.

Here is the sh script：

MEM_LIMITE=20gb
HTTP_PORT=80
HTTPS_PORT=443
docker create --runtime=nvidia --gpus all --restart always -p $HTTPS_PORT:80 -p $HTTP_PORT:443 --ulimit memlock=819200000:819200000 -m $MEM_LIMITE --memory-swap $MEM_LIMITE -w /opt/idfort --entrypoint "/opt/idfort/bin/idfort" --name idfort idfort/gpu:20240603 -container 1

Once containers are created, a developer can list the containers using the following command.

$ docker ps -a

This command shall retrieve a list of all available containers including the one that is generated against ID Fort Docker Image.

CONTAINER ID   IMAGE           COMMAND                  CREATED          STATUS   PORTS   NAMES
4a45fa8fe6c2   idfort:20240531 "/opt/idfort/bin/idf…"   22 seconds ago   Created          idfort

Start the container by using the below command. The developer needs to ensure the CONTAINER ID as as per the above list.

$ docker start 4a45fa8fe6c2

Or

$ docker start idfort

e) Administrator Setup

When the container runs for the first time, you must setup an administrator credential. To do so, open your browser and visit.

http://yourserverip/portal

You will be prompted to create an administrator account.

f) License Registration

After creating an admin account, login the portal as an admin and visit the Server -> License Status page. Fill in your license certificate and click Register.

After registering, you should perform an update using Server -> Server Status page to install all the extra modules that comes with your purchased license.

g) Migrating License

Once a license is registered, it is locked into the hardware and cannot be registered again. If you plan on moving the software to a different server, you have two options:

1. Use docker export to save the container into a .tar file, then docker load on a new machine. This method will retain all data and licenses. However, you must stop and delete the original container to prevent conflict in license check.
2. Under Admin Portal > Server > License Status, click on "Deactivate License" button to deactivate the current license; this allows you to register your license certificate on a new machine.