Installation & Setup

This section provides guide on how to install and run the AI Verify Toolkit.

Run using Docker Compose

Refer to https://github.com/aiverify-foundation/aiverify/tree/v2.x/deployment/docker-compose for detailed installation instructions.

Prerequisites

• Host OS: Linux and MacOS
• Docker

Step 1: Download the Docker Compose File

Download the compose.yaml file from the official repository.

Step 2: Install and Start Services

Mode 1 (Manual Test Execution and Upload)

Run the following command to install the portal and related services:

docker-compose --profile portal up -d  

Mode 2 (Automated Test Execution via Portal)

Run the following command to install the portal, test engine worker, and related services:

docker-compose --profile portal --profile automated-tests-venv up -d  

After installation, you can access the portal at http://localhost:3000.

Run using Kubernetes

Refer to https://github.com/aiverify-foundation/aiverify/tree/v2.x/deployment/kubernetes for detailed installation instructions.

Note: The provided YAML files are for reference and may need to be adapted to your specific environment. Consider the following:

• Resource Limits: Adjust CPU and memory requests/limits based on your cluster's capacity.
• Storage: Configure persistent storage if required.
• Networking: Modify service types (e.g., ClusterIP, NodePort, LoadBalancer) based on your networking setup.
• Security: Implement security best practices, such as using secrets for sensitive data and enabling network policies.

Step 1: Create the aiverify Namespace

kubectl create namespace aiverify

Step 2: Deploy AI Verify Using Kubernetes YAML Files

1. Sparse checkout the K8s YAML files: Sparse Checkout of Kubernetes Deployment Files

   To clone only the deployment/kubernetes folder from the v2.x branch of the repository, use the following commands:

   # Clone the repository with sparse checkout for the specific folder
   git clone --filter=blob:none --sparse -b v2.x https://github.com/aiverify-foundation/aiverify.git
   cd aiverify
   git sparse-checkout set deployment/kubernetes

2. Apply the YAML Files: Apply the YAML files to deploy AI Verify in the aiverify namespace.

   kubectl apply -f <yaml-file> -n aiverify

   Repeat this command for each YAML file (e.g., apigw.yaml, portal.yaml, etc.).

   kubectl apply -f apigw.yaml
   kubectl apply -f portal.yaml
   kubectl apply -f valkey.yaml
   kubectl apply -f test_engine_worker.yaml

Step 3: Verify the Deployment

After deploying AI Verify, verify that all components are running correctly.

1. Check Pods: Ensure that all pods are in the Running state.

   kubectl get pods -n aiverify

2. Check Services: Verify that the services are correctly exposed.

   kubectl get svc -n aiverify

3. Access AI Verify: Depending on your setup, access AI Verify via the exposed service (e.g., using the service's external IP or port-forwarding).

   kubectl port-forward svc/portal -n aiverify 3000:3000 & \
   kubectl port-forward svc/apigw -n aiverify 4000:4000 &

Then, open your browser and navigate to http://localhost:3000.