Deployment - Azure VM

September 8, 2023 · View on GitHub

Prerequisites

Azure subscription
Subscription access to Azure OpenAI service. Start here to Request Access to Azure OpenAI Service

Deployment steps

Follow the steps below to deploy the solution to your Azure subscription.

Run the following script to provision a development VM with Visual Studio 2022 Community and required dependencies preinstalled.
```
.\scripts\Deploy-Vm.ps1 -resourceGroup <rg_name> -location <location> -password <password>
```
<password> is the password for the BYDtoChatGPTUser account that will be created on the VM. It must be at least 12 characters long and meet the complexity requirements of Azure VMs.

When the script completes, the console output should display the name of the provisioned VM similar to the following:
```
The resource prefix used in deployment is libxarwttxjde
The deployed VM name used in deployment is libxarwttxjdevm
```
Use RDP to remote into the freshly provisioned VM with the username BYDtoChatGPTUser and the password you provided earlier on.
Add the BYDtoChatGPTUser account to the docker-users local group on the VM. Sign out and sign back in to the VM to apply the changes.
Install WSL2 by running the following command in a command prompt:
```
wsl --install
```
Restart the VM to complete the setup.
Log back in with the BYDtoChatGPTUser account and start Docker Desktop. Ensure the Docker engine is up and running. Keep Docker Desktop running in the background.

Clone the repository:

git clone https://github.com/AzureCosmosDB/VectorSearchAiAssistant.git

Switch to the cognitive-search-vector branch:

cd VectorSearchAiAssistant
git checkout cognitive-search-vector

Run the following script to provision the infrastructure and deploy the API and frontend. This will provision all of the required infrastructure, deploy the API and web app services into AKS, and import data into Cosmos DB.
```
./scripts/Unified-Deploy.ps1 -resourceGroup <rg_name> -location <location> -subscription <target_subscription_id>
```

NOTE: Make sure to set the <location> value to a region that supports Azure OpenAI services. See Azure OpenAI service regions for more information.