On-Premise Installation Guide for Cognipeer

This guide explains how to deploy the Cognipeer application on-premise using a single Docker container. With the proper configuration, the application will be ready to use immediately. A scalable version is planned for future release.

Note: The Cognipeer image is distributed via AWS ECR. You will need the provided accessKey and Secret credentials from your company to access the repository.

Prerequisites

• Docker: Ensure Docker is installed on your host machine.
• AWS CLI: Installed and configured with your provided AWS credentials to authenticate with ECR.
• (Optional) Kubernetes: If you plan to deploy Cognipeer on a Kubernetes cluster.

Environment Variables

Before starting the container, define the following environment variables:

Variable	Description	Default Value
LICENCE	Your application license key.	Required
ADMIN_EMAIL	Administrator email address.	Required
ADMIN_PASSWORD	Administrator password.	Required
EMAIL_SERVER	Email server for sending notifications.	Optional
EMAIL_PORT	Port for the email server.	Optional
EMAIL_USERNAME	Username for the email server.	Optional
EMAIL_PASSWORD	Password for the email server.	Optional
EMAIL_FROM	Default sender email address.	Optional
LOCAL_FILE_PATH	Path for local file storage.	Optional ./files
LOCAL_VECTOR_DB_PATH	Path for the local vector database.	Optional ./data/vector
MONGODB_URL	URL for the MongoDB instance.	Optional Will use local MongoDB instance if empty
APPLICATION_URL	URL for the application.	Optional http://127.0.0.1:3000
API_URL	URL for the API endpoint.	Optional http://127.0.0.1:8080

Once these are set, the application will run with the default configuration.

Ports

Port	Description
8080	API
3000	UI
27017	MongoDB

Paths for Volume Mapping

Description	Default Path
Local vector database	/app/api-source/data
Local file storage	/app/api-source/files
MongoDB Files	/data/db

Docker Deployment

Step 1: Log in to AWS ECR

Use the AWS CLI to log in to your ECR repository:

aws ecr get-login-password --region <region> --profile <profile> | docker login --username AWS --password-stdin <aws_account_id>.dkr.ecr.<region>.amazonaws.com

Replace 'region' and 'aws_account_id' with your actual AWS region and account ID.

Step 2: Run the Docker Container

Run the container with the necessary environment variables:

docker run -d \
  -e LICENCE="<your-licence-key>" \
  -e ADMIN_EMAIL="<your-admin-email>" \
  -e ADMIN_PASSWORD="<your-admin-password>" \
  -e EMAIL_SERVER="<optional-email-server>" \
  -e EMAIL_PORT="<optional-email-port>" \
  -e EMAIL_USERNAME="<optional-email-username>" \
  -e EMAIL_PASSWORD="<optional-email-password>" \
  -e EMAIL_FROM="<optional-email-from>" \
  -e LOCAL_FILE_PATH="./files" \
  -e LOCAL_VECTOR_DB_PATH="./data/vector" \
  -e MONGODB_URL="<mongodb-url>" \
  -e APPLICATION_URL="<application-url>" \
  -e API_URL="<api-url>" \
  <aws_account_id>.dkr.ecr.<region>.amazonaws.com/cognipeer:latest

Be sure to replace the placeholder values with your actual configuration details.

Kubernetes Deployment

If you prefer to deploy Cognipeer using Kubernetes, use the following sample YAML as a starting point. Adjust the values as needed for your environment.

apiVersion: apps/v1
kind: Deployment
metadata:
  name: cognipeer
spec:
  replicas: 1
  selector:
    matchLabels:
      app: cognipeer
  template:
    metadata:
      labels:
        app: cognipeer
    spec:
      containers:
        - name: cognipeer
          image: <aws_account_id>.dkr.ecr.<region>.amazonaws.com/cognipeer:latest
          env:
            - name: LICENCE
              value: "<your-licence-key>"
            - name: ADMIN_EMAIL
              value: "<your-admin-email>"
            - name: ADMIN_PASSWORD
              value: "<your-admin-password>"
            - name: EMAIL_SERVER
              value: "<optional-email-server>"
            - name: EMAIL_PORT
              value: "<optional-email-port>"
            - name: EMAIL_USERNAME
              value: "<optional-email-username>"
            - name: EMAIL_PASSWORD
              value: "<optional-email-password>"
            - name: EMAIL_FROM
              value: "<optional-email-from>"
            - name: LOCAL_FILE_PATH
              value: "./files"
            - name: LOCAL_VECTOR_DB_PATH
              value: "./data/vector"
            - name: MONGODB_URL
              value: "<mongodb-url>"
            - name: APPLICATION_URL
              value: "<application-url>"
            - name: API_URL
              value: "<api-url>"

Apply the deployment using:

kubectl apply -f cognipeer-deployment.yaml

Ensure your Kubernetes cluster is properly configured to pull images from AWS ECR.

Post-Installation Configuration

Once the application is running, complete the following steps within the Cognipeer interface:

Add a Model

Navigate to the model management section and add a new model to the system.

Configure Workspace Settings

• Go to the settings/workspace section.
• Select the default model.
• Enter the embedding model details as required.

After these configurations, the system will be fully operational.

Conclusion

You have now deployed Cognipeer on-premise using Docker (or Kubernetes) and completed the necessary post-installation setup. If you encounter any issues or need further assistance, please refer to our contact support.