Skip to content

Basic Installation

Distribution channels and supported tags

We distribute our container images via the following container registries:

Images are RedHat Certified and fully identical. The face-api image is based on the official ubi9-minimal image, maintained by RedHat. The image is configured to be run as an arbitrarily assigned UID (OpenShift-ready).

The supported tags are the following:

  • nightly-cpu/nightly-gpu - nightly build for development purposes, it should never be used in production
  • A.B.C.D-develop-cpu/A.B.C.D-develop-gpu - release candidate build
  • A.B.C.D-cpu/A.B.C.D-gpu - stable release build

For more information about versioning of FaceAPI SDK, please click here.

Warning

To obtain the latest SDK / document database / service wrapper version, please pull the container image with the newest TAG version according to the above-mentioned description.

Windows 10/11

  1. Install Docker Desktop by following the instructions based on your Windows edition:

    Warning

    Docker Desktop doesn't support Windows Server version installation.

  2. Start the Docker Desktop application.

  3. Put your license to your Desktop directory.
  4. Open the Command line and execute the following command:

    docker run -d -p 41101:41101 -v C:\%HOMEPATH%\Desktop\regula.license:/app/extBin/unix_x64/cpu/regula.license regulaforensics/face-api:latest
    

    Info

    To open the Command line, press the WIN button, type 'cmd', and press Enter.

  5. To make sure that the service is up and running, enter the http://localhost:41101/ address in a web browser.

Linux and macOS

  1. Install Docker Engine:

  2. Put your license to your user's home (/home/username/) directory.

  3. Run Container

    CPU version

    Open a terminal and execute:

    docker run -p 41101:41101 -v ~/regula.license:/app/extBin/unix_x64/cpu/regula.license regulaforensics/face-api:latest
    

    GPU version

    Warning

    GPU version installation requires video card on board and nvidia drivers installed. For more information, please visit the official docker documentation.

    Open a terminal and execute:

    docker run -p 41101:41101 --gpus all -v ~/regula.license:/app/extBin/unix_x64/gpu/regula.license regulaforensics/face-api:3.2.325.304-gpu
    
  4. To make sure that the service is up and running, enter the http://localhost:41101/ address in a web browser.

Regular

  1. Create the namespace "regula":

    kubectl create namespace regula
    

  2. Import the secret "regula.license":

    kubectl create secret generic regula-license -n regula --from-file=regula.license
    

  3. Download the deployment file and run:

    kubectl apply -f k8sface-api.yml
    

GPU

Info

GPU version requires nvidia gpu nodes. For more information please visit https://github.com/NVIDIA/k8s-device-plugin

  1. Create the namespace "regula":

    kubectl create namespace regula
    

  2. Import the secret "regula.license":

    kubectl create secret generic regula-license -n regula --from-file=regula.license
    

  3. Download the deployment file and run:

    kubectl apply -f k8sface-api-gpu.yml
    

HTTPS via Nginx

  1. Create the namespace "regula":

    kubectl create namespace regula
    

  2. Import the secret "regula.license":

    kubectl create secret generic regula-license -n regula --from-file=regula.license
    

  3. Create ConfigMap "default" from nginx config file:

    kubectl create configmap default -n regula --from-file=default.conf
    

  4. Create a TLS secret that contains a TLS private key and certificate:

    kubectl create secret tls certificates -n regula --key="tls.key" --cert="tls.crt"
    

  5. Download the deployment file and run:

    kubectl apply -f k8sface-api-https.yml
    

Add Repo

helm repo add regulaforensics https://regulaforensics.github.io/helm-charts
helm repo update

See helm repo for command documentation.

Prerequisites

  • At least 3 GB of RAM available on your cluster per pod's worker
  • Helm 3
  • PV provisioner support in the underlying infrastructure (essential for storing logs)

Installing the Chart

To install the chart with the release name my-release:

helm install my-release regulaforensics/faceapi

Uninstalling the Chart

To uninstall/delete the my-release deployment:

helm uninstall my-release

The command removes all the Kubernetes components associated with the chart and deletes the release.

Settings

Helm Chart settings can be found at helm repo.

If you want to change multiple configurations, you can download the values.yaml file.