Deploy the Azure Search Backup and Restore application in an Azure Function

This project describes the steps for deploying the Azure Search backup and restore application within an Azure Function.

Azure Functions is a serverless compute service that lets you run code without having to explicitly provision or manage infrastructure. Azure Functions uses a Pay-per-use pricing model, meaning customers only have to pay for the duration of time the Function runs. Furthermore, Azure Functions supports per-second billing.

This Function is configured with a Timer input trigger which can be configured to periodically synchronize the knowledge base documents between the Source and Target Azure Search instances.

Follow the steps below for building and running the Function application on Azure.

1. (Optional) Fork this GitHub repository.

   If you intend to modify the Function code before deploying it to Azure, you will need to Fork this repository into your GitHub account. Using a browser, login to your GitHub account and Fork this repository.

2. Open an Azure Cloud Shell session.

   We will clone this repository, build and deploy the Azure Function on Azure via the Azure Cloud Shell. Using a web brower, access the [Azure Cloud Shell] (https://shell.azure.com/) and open a Linux Bash session.

3. Create a Function Application in Azure.

   Refer to the Azure Function documentation to create an Function Application. Alternatively, you can also create the Function application via the Azure Portal.

4. Update the Azure Function configuration parameters.

   Using a web browser, access the Azure Function application which you just created. Access the Configuration panel and create the application parameters with corresponding values as per the table below.

   Function Parameter Name	Description
   AZURE_STORAGE_CONNECTION_STRING	Specify the Connection String to an Azure Storage account.
   AzStorageContainerName	Specify a storage Container name in the Azure Storage account specified by the parameter 'AZURE_STORAGE_CONNECTION_STRING'. This container will be used to store the knowledge base documents from the source Azure Search instance and then restore it in the target instance.
   SrcSearchSvcName	Specify the source Azure Search instance Service Name.
   TgtSearchSvcName	Specify the target Azure Search instance Service Name.
   SrcSearchApiKey	Specify the source Azure Search instance API Key.
   TgtSearchApiKey	Specify the target Azure Search instance API Key.

5. Clone this GitHub repository.

   In the Azure cloud shell, clone this GitHub repository to the git-repos directory. Refer to the command snippet below.

   # Create the root directory to store all your projects.
   $ mkdir git-repos
   #
   # Switch current directory to 'git-repos'
   $ cd git-repos
   #
   # Clone the master or your GitHub repository. Cloning from your GitHub account 
   # will allow you to make changes to the application artifacts.
   # Substitute your GitHub Account ID in the URL as below.
   $ git clone https://github.com/<YOUR-GITHUB-ACCOUNT/QnAMakerBackupRestore.git
   #

6. (Optional) Update the Function Input Trigger.

   A Timer Trigger lets you run a Function on a schedule.

   This Function is configured to trigger @ 08:30am during the weekdays (Mon. - Fri. by default). To configure a different time interval for synchronizing the knowledge bases in the source and target Azure Search instances, modify the Function's Input Trigger. Edit the ./AzSearch/AzSearchBackupRestore.cs class and update the TimerTrigger annotation value in the main Function method Run.

   Azure Function uses the NCronTab library to interpret NCRONTAB expressions. To configure the timer interval, refer to the Azure Function documentation here.

7. Deploy the Function on the Azure Function Application.

   Refer to the commands in the snippet below to deploy the Function on Azure. Substitute the correct name of your Azure Function Application in the command.

   # Use Azure Function Core Tools to deploy the Function on Azure.
   # Substitute the correct value for the Function Application name.
   #
   $ func azure functionapp publish <name-of-function-application>
   #

8. Confirm and verify the Function deployment on Azure.

   Use the Azure portal to verify the Azure Function got deployed on Azure.

   Verify the Function runs as per the configured schedule. Use the Function Monitor tab to view the messages logged by the application.