Installing QAnswer on premise

Introduction

If you are interested to install QAnswer on premise you need to reach us at info@the-qa-company.com to get a license key and the various access to be able to follow this guide.

Architecture of the application

Here is a diagram explaining the overall architecture of the qanswer application

QAnswer contains the following software components:

• a React3 front-end web-app to allow access and interact with APIs via an intuitive user interface;
• a Spring Boot Java back-end4. This application is serving and securing all APIs;
• a fast API python back-end5 that is responsible for the machine learning functionalities;
• a fast API python back-end that is serving a language model;
• an ElasticSearch instance6;
• a Redis7 server for caching;
• a Selenium8 server to scrape websites;
• a transcription server for speech to text;
• a Postgres9 database that stores user information.

Requirements

To run QAnswer you need:

• a linux distribution (we use Ubuntu by default)
• A GPU with at least 48GB of memory (we recommend NVIDIA® L40S)
• Docker

(this will allow you to serve 7 to 13 concurrent users depening on RAG or Chat workload, you can contact us at info@the-qa-company.com if you need help sizing your infrastructure)

Install Nvidia Drivers

1. Download and install the CUDA drivers from the NVIDIA website. If you are using a Ubuntu distribution you can use the following command to install the drivers

sudo ubuntu-drivers install --gpgpu

Install Docker

You need to install Docker on your machine. You can follow the official documentation to install it on your machine.

Then you need to install the nvidia container runtime by following the official documentation

After that you need to restart docker by running the following command

sudo service docker restart

Install QAnswer

QAnswer is provided as a docker-compose file. You can clone the repository by running the following command

git clone https://gitlab.the-qa-company.com/qanswer-app/bundle.git

The first thing you have to do is from template.env create a .env file with the secret of your installation. In this file you have to fill the following field:

Configuration overview

System details

The credential used to get the images of QAnswer (supplied by us)

Property	Value
USER	The username to access the registry
PASSWORD	The password to access the registry

Application details

Property	Value
QA_KEY	The license Key to activate the application we provide
QA_PUBLIC_UI_URL	The URL of the frontend
QA_PUBLIC_URL	The URL of the backend for http/https
QA_PUBLIC_WSURL	The URL of the backend for websocket
QA_ADMIN_USERNAME	The username for the admin account
QA_ADMIN_EMAIL	The email for the admin account
QA_ADMIN_PASSWORD	the password for the admin account
SPRING_DATASOURCE_PASSWORD	The password of the database
APP_JWTSECRET	A JWT secret token
CHECK_MALWARE	True to enable Malware detection else False

Malware detection

Malware detection in ClamAV (Clam AntiVirus) refers to the process of identifying malicious software on systems using a signature-based and heuristic scanning engine. ClamAV is an open-source antivirus engine widely used for email scanning, web scanning, and endpoint security. It detects malware by comparing files against a database of known virus signatures—unique patterns associated with specific types of malware. This signature database is regularly updated to cover new threats. ClamAV also incorporates heuristic detection, which involves analyzing files for suspicious behaviors that may indicate a previously unknown threat. Through these mechanisms, ClamAV can identify and quarantine or remove various types of malware, including viruses, trojans, and ransomware.

note

If this service is enabled and a file is marked as malware by ClamAV, the file upload is skipped.

warning

To enable malware detection add --profile malwareDetection to the docker compose.

Contact Email

The email that people will be encourage to contact in case there is a problem

Property	Value
CONTACT_MAIL	The email of the contact

Run QAnswer

Then after filling everything you have to run to generate the configuration of QAnswer

./generate-config.sh

After that, pull all the image of QAnswer by running the following command

./init.sh

Then you can run finally the following command to have QAnswer running

docker compose up -d

If you want to enable malware detection add --profile malwareDetection

docker compose --profile malwareDetection up -d

warning

After enabling the malwareDetection profile you have to add the env variable CHECK_MALWARE to the env file to start using the service.