Installation Guide

This guide walks you through deploying a Clariti Connector from start to finish. The entire process takes about 10 minutes.

Prerequisites

Before you begin, make sure you have:

• Docker Engine 24+ (or Podman 4+) installed on a Linux host
• Outbound HTTPS access to *.clariti-ai.xyz (port 443)
• Network access to the on-premises tools you want to monitor (e.g., Active Directory on port 636, SIEM API)
• Minimum resources: 1 CPU, 512 MB RAM, 1 GB disk
• Clariti account with Admin role

Step 1: Create a Connector in Clariti

1. Log in to your Clariti dashboard
2. Navigate to Connectors in the left sidebar
3. Click Add Connector
4. Enter a name (e.g., "Denver Office Connector")
5. Select or create a Site (a site groups connectors by physical location)
6. Click Create Connector
7. Copy the registration token — you'll need it in Step 3. This token expires in 24 hours and can only be used once.

Step 2: Deploy the Docker Container

Option A: Docker Run (Quick Start)

docker run -d \
  --name clariti-connector \
  -e CLARITI_CLOUD_BASE_URL=https://api.clariti-ai.xyz \
  -e CLARITI_LOG_LEVEL=info \
  -p 8443:8443 \
  -v connector-data:/data \
  --read-only \
  --tmpfs /tmp \
  --memory=512m \
  --cpus=1 \
  --cap-drop=ALL \
  --security-opt=no-new-privileges \
  --restart unless-stopped \
  ghcr.io/clariti-ai/connector:latest

Option B: Docker Compose (Recommended)

Create a docker-compose.yml file:

services:
  connector:
    image: ghcr.io/clariti-ai/connector:latest
    container_name: clariti-connector
    read_only: true
    tmpfs:
      - /tmp
    volumes:
      - connector-data:/data
    environment:
      - CLARITI_CLOUD_BASE_URL=https://api.clariti-ai.xyz
      - CLARITI_LOG_LEVEL=info
    ports:
      - "8443:8443"
    deploy:
      resources:
        limits:
          memory: 512M
          cpus: '1'
    security_opt:
      - no-new-privileges:true
    cap_drop:
      - ALL
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "clariti-connector", "status"]
      interval: 30s
      timeout: 10s
      retries: 3

volumes:
  connector-data:

Then run:

docker compose up -d

Step 3: Register the Connector

1. Open https://<connector-ip>:8443 in your browser
2. You'll see a self-signed certificate warning — this is expected for the registration UI. Accept it to proceed.
3. Enter the registration token you copied in Step 1
4. Click Register
5. The connector will contact Clariti, exchange credentials, and establish the secure link

After successful registration, port 8443 is no longer needed. You can remove the -p 8443:8443 port mapping for additional security.

Step 4: Verify in the Dashboard

1. Go back to your Clariti dashboard
2. Navigate to Connectors
3. Your connector should show as Online within 60 seconds
4. Click into the connector to see health metrics (CPU, memory, disk, uptime)

Step 5: Configure Adapters

With the connector online, you can now configure adapters for your on-premises tools:

1. Go to Adapters in the dashboard
2. Click Add Adapter and select an on-premises tool
3. Enter the credentials (these are encrypted and sent securely to the connector)
4. Trigger a first scan to verify connectivity

Environment Variables

Persistent Storage

The connector stores buffered data in /data inside the container. Mount this as a Docker volume to preserve data across restarts:

• Credentials are encrypted at rest and persisted during graceful shutdown
• Buffered results are stored in SQLite if the cloud is temporarily unreachable
• Buffered logs are prioritized — error and warning logs are preserved longer than debug logs

Updating the Connector

The connector supports three update policies (configurable in the dashboard):

• Manual — you pull the new image and restart manually
• Delayed (default) — the connector auto-updates 24 hours after a new version is released
• Auto — the connector updates immediately when a new version is available

To manually update:

docker pull ghcr.io/clariti-ai/connector:latest
docker compose down && docker compose up -d