Running brighter Redact Enterprise (Three Container Mode)

For latest documentation please refer to redact-enterprise-orchestration on github.

!!! ---- DEPRECATED GUIDE BELOW ---- !!!

brighter Redact Enterprise

brighter Redact Enterprise is an ecosystem comprising of three containers, which should be started separately. The three containers - redact, redact-gpu and redact-utils are described in the architecture below.


Ecosystem of brighter Redact Enterprise

Running brighter Redact services


  1. Log in to the brighter AI docker registry with your credentials (only needed once per machine):
    docker login
  2. Make sure you have access to your redact license file. For this guide, we'll assume that it's stored in /home/dev/license.bal


Usage-based licenses

If you're using a usage-based license you must have an active internet connection at all times!

Starting brighter Redact Enterprise

  1. Start the redact-gpu image on all gpus by running:
    docker run -d --gpus all -p 8000:8000 -p 8001:8001 -p 8003:8003


Docker image tag

We highly recommend pinning your tag to a specific version for production systems.

You can verify that this has started by looking for redact-gpu in the output generated by running the command:
docker container ls

  1. Wait 5 minutes for redact-gpu to come fully online before running the next command. You can check the health endpoint to verify that the container is ready with the following command:

curl -v $HOSTIP:8000/v2/health/ready

The HTTP request should return status 200. If the response is not 200, it means that redact-gpu is not ready yet.


Update $HOST_IP in commands below.

Please ensure with all instances of $HOSTIP in the commands are replaced with the IP of the computer you are hosting the container on.

  1. Start the main redact container by running the following command.

docker run -e REDACT_GPU_URL=$HOSTIP:8001 -e LP_DETECTION_SERVICE=$HOSTIP:8003 -p 8787:8787 -t -d -v /home/dev/license.bal:/license.bal

  1. Make sure that the main container has started correctly by running the command
    curl -X GET "http://$HOSTIP:8787/services/v3/status" -H "accept: application/json"

This should return:
{ "status": "operational"}

If it fails or returns:
{"status": "not_operational" }

Please make sure you have a valid license and have followed the steps 1-5 correctly.

Starting brighter Redact User Interface (UI)

  1. Next start the redact-utils container by running the following command with all instances of $HOSTIP replaced with the ip of the computer you are hosting the redact container on:
    docker run --rm -d -p 8080:80 -e REDACT_URL=http://$HOSTIP:8787

  2. Make sure that the redact-utils container has started correctly by checking that either the user interface for anonymization ($HOSTIP:8080/ui) or the selective redaction application ($HOSTIP:8080/sra) is up.

  3. Start anonymizing using the ui ($HOSTIP:8080/ui), sra ($HOSTIP:8080/sra), or flassger interface($HOSTIP:8787).

Limiting GPUs

Simply change the --gpus flag to select your desired GPUs. E.g.:
docker run -d --gpus '"device=1,2"' -p 8001:8001 -p 8003:8003
See here for more information.

API reference

You can easily access the API documentation under: