Tutorial: Create a write-behind cache by using Azure Functions and Azure Cache for Redis

  • Article

The objective of this tutorial is to use an Azure Cache for Redis instance as a write-behind cache. The write-behind pattern in this tutorial shows how writes to the cache trigger corresponding writes to a SQL database (an instance of the Azure SQL Database service).

You use the Redis trigger for Azure Functions to implement this functionality. In this scenario, you see how to use Azure Cache for Redis to store inventory and pricing information, while backing up that information in a SQL database.

Every new item or new price written to the cache is then reflected in a SQL table in the database.

In this tutorial, you learn how to:

  • Configure a database, trigger, and connection strings.
  • Validate that triggers are working.
  • Deploy code to a function app.

Prerequisites

  • An Azure subscription. If you don't have an Azure subscription, create a Trial.
  • Completion of the previous tutorial, Get started with Azure Functions triggers in Azure Cache for Redis, with these resources provisioned:
    • Azure Cache for Redis instance
    • Azure Functions instance
    • A working knowledge of using Azure SQL
    • Visual Studio Code (VS Code) environment set up with NuGet packages installed

Create and configure a new SQL database

The SQL database is the backing database for this example. You can create a SQL database through the Azure portal or through your preferred method of automation.

For more information on creating a SQL database, see Quickstart: Create a single database - Azure SQL Database.

This example uses the portal:

  1. Enter a database name and select Create new to create a new server to hold the database.

    Screenshot of creating an Azure SQL resource.

  2. Select Use SQL authentication and enter an admin sign-in and password. Be sure to remember these credentials or write them down. When you're deploying a server in production, use Microsoft Entra authentication instead.

    Screenshot of the authentication information for an Azure SQL resource.

  3. Go to the Networking tab and choose Public endpoint as a connection method. Select Yes for both firewall rules that appear. This endpoint allows access from your Azure function app.

    Screenshot of the networking setting for an Azure SQL resource.

  4. After validation finishes, select Review + create and then Create. The SQL database starts to deploy.

  5. After deployment finishes, go to the resource in the Azure portal and select the Query editor tab. Create a new table called inventory that holds the data you'll write to it. Use the following SQL command to make a new table with two fields:

    • ItemName lists the name of each item.
    • Price stores the price of the item.
    CREATE TABLE inventory (
    ItemName varchar(255),
    Price decimal(18,2)
    );

    Screenshot showing the creation of a table in Query Editor of an Azure SQL resource.

  6. After the command finishes running, expand the Tables folder and verify that the new table was created.

Configure the Redis trigger

First, make a copy of the same VS Code project that you used in the previous tutorial. Copy the folder from the previous tutorial under a new name, such as RedisWriteBehindTrigger, and open it in VS Code.

Second, delete the RedisBindings.cs and RedisTriggers.cs files.

In this example, you use the pub/sub trigger to trigger on keyevent notifications. The goals of the example are:

  • Trigger every time a SET event occurs. A SET event happens when either new keys are written to the cache instance or the value of a key is changed.
  • After a SET event is triggered, access the cache instance to find the value of the new key.
  • Determine if the key already exists in the inventory table in the SQL database.
    • If so, update the value of that key.
    • If not, write a new row with the key and its value.

To configure the trigger:

  1. Import the System.Data.SqlClient NuGet package to enable communication with the SQL database. Go to the VS Code terminal and use the following command:

      dotnet add package System.Data.SqlClient

  2. Create a new file called RedisFunction.cs. Make sure you've deleted the RedisBindings.cs and RedisTriggers.cs files.

  3. Copy and paste the following code in RedisFunction.cs to replace the existing code:

using Microsoft.Extensions.Logging;
using Microsoft.Azure.Functions.Worker;
using Microsoft.Azure.Functions.Worker.Extensions.Redis;
using System.Data.SqlClient;

public class WriteBehindDemo
{
    private readonly ILogger<WriteBehindDemo> logger;

    public WriteBehindDemo(ILogger<WriteBehindDemo> logger)
    {
        this.logger = logger;
    }
    
    public string SQLAddress = System.Environment.GetEnvironmentVariable("SQLConnectionString");

    //This example uses the PubSub trigger to listen to key events on the 'set' operation. A Redis Input binding is used to get the value of the key being set.
    [Function("WriteBehind")]
    public void WriteBehind(
        [RedisPubSubTrigger(Common.connectionString, "__keyevent@0__:set")] Common.ChannelMessage channelMessage,
        [RedisInput(Common.connectionString, "GET {Message}")] string setValue)
    {
        var key = channelMessage.Message; //The name of the key that was set
        var value = 0.0;

        //Check if the value is a number. If not, log an error and return.
        if (double.TryParse(setValue, out double result))
        {
            value = result; //The value that was set. (i.e. the price.)
            logger.LogInformation($"Key '{channelMessage.Message}' was set to value '{value}'");
        }
        else
        {
            logger.LogInformation($"Invalid input for key '{key}'. A number is expected.");
            return;
        }        

        // Define the name of the table you created and the column names.
        String tableName = "dbo.inventory";
        String column1Value = "ItemName";
        String column2Value = "Price";        
        
        logger.LogInformation($" '{SQLAddress}'");
        using (SqlConnection connection = new SqlConnection(SQLAddress))
            {
                connection.Open();
                using (SqlCommand command = new SqlCommand())
                {
                    command.Connection = connection;

                    //Form the SQL query to update the database. In practice, you would want to use a parameterized query to prevent SQL injection attacks.
                    //An example query would be something like "UPDATE dbo.inventory SET Price = 1.75 WHERE ItemName = 'Apple'".
                    command.CommandText = "UPDATE " + tableName + " SET " + column2Value + " = " + value + " WHERE " + column1Value + " = '" + key + "'";
                    int rowsAffected = command.ExecuteNonQuery(); //The query execution returns the number of rows affected by the query. If the key doesn't exist, it will return 0.

                    if (rowsAffected == 0) //If key doesn't exist, add it to the database
                 {
                         //Form the SQL query to update the database. In practice, you would want to use a parameterized query to prevent SQL injection attacks.
                         //An example query would be something like "INSERT INTO dbo.inventory (ItemName, Price) VALUES ('Bread', '2.55')".
                        command.CommandText = "INSERT INTO " + tableName + " (" + column1Value + ", " + column2Value + ") VALUES ('" + key + "', '" + value + "')";
                        command.ExecuteNonQuery();

                        logger.LogInformation($"Item " + key + " has been added to the database with price " + value + "");
                    }

                    else {
                        logger.LogInformation($"Item " + key + " has been updated to price " + value + "");
                    }
                }
                connection.Close();
            }

            //Log the time that the function was executed.
            logger.LogInformation($"C# Redis trigger function executed at: {DateTime.Now}");
    }
}

Important

This example is simplified for the tutorial. For production use, we recommend that you use parameterized SQL queries to prevent SQL injection attacks.

Configure connection strings

You need to update the local.settings.json file to include the connection string for your SQL database. Add an entry in the Values section for SQLConnectionString. Your file should look like this example:

{
  "IsEncrypted": false,
  "Values": {
    "AzureWebJobsStorage": "",
    "FUNCTIONS_WORKER_RUNTIME": "dotnet-isolated",
    "redisConnectionString": "<redis-connection-string>",
    "SQLConnectionString": "<sql-connection-string>"
  }
}

To find the Redis connection string, go to the resource menu in the Azure Cache for Redis resource. Locate the string is in the Access Keys area on the Resource menu.

To find the SQL database connection string, go to the resource menu in the SQL database resource. Under Settings, select Connection strings, and then select the ADO.NET tab. The string is in the ADO.NET (SQL authentication) area.

You need to manually enter the password for your SQL database connection string, because the password isn't pasted automatically.

Build and run the project

  1. In VS Code, go to the Run and debug tab and run the project.

  2. Go back to your Azure Cache for Redis instance in the Azure portal, and select the Console button to enter the Redis console. Try using some SET commands:

    • SET apple 5.25
    • SET bread 2.25
    • SET apple 4.50

  3. Back in VS Code, the triggers are being registered. To validate that the triggers are working:

    1. Go to the SQL database in the Azure portal.

    2. On the resource menu, select Query editor.

    3. For New Query, create a query with the following SQL command to view the top 100 items in the inventory table:

      SELECT TOP (100) * FROM [dbo].[inventory]

      Confirm that the items written to your Azure Cache for Redis instance appear here.

    Screenshot showing the information has been copied to SQL from the cache instance.

Deploy the code to your function app

This tutorial builds on the previous tutorial. For more information, see Deploy code to an Azure function.

  1. In VS Code, go to the Azure tab.

  2. Find your subscription and expand it. Then, find the Function App section and expand that.

  3. Select and hold (or right-click) your function app, and then select Deploy to Function App.

Add connection string information

This tutorial builds on the previous tutorial. For more information on the redisConnectionString, see Add connection string information.

  1. Go to your function app in the Azure portal. On the resource menu, select Environment variables.

  2. In the App Settings pane, enter SQLConnectionString as a new field. For Value, enter your connection string.

  3. Select Apply.

  4. Go to the Overview blade and select Restart to restart the app with the new connection string information.

Verify deployment

After the deployment finishes, go back to your Azure Cache for Redis instance and use SET commands to write more values. Confirm that they also appear in your SQL database.

If you want to confirm that your function app is working properly, go to the app in the portal and select Log stream from the resource menu. You should see the triggers running there, and the corresponding updates being made to your SQL database.

If you ever want to clear the SQL database table without deleting it, you can use the following SQL query:

TRUNCATE TABLE [dbo].[inventory]

Clean up resources

If you want to continue to use the resources you created in this article, keep the resource group.

Otherwise, to avoid charges related to the resources, if you're finished using the resources, you can delete the Azure resource group that you created.

Warning

Deleting a resource group is irreversible. When you delete a resource group, all the resources in the resource group are permanently deleted. Make sure that you do not accidentally delete the wrong resource group or resources. If you created the resources inside an existing resource group that has resources you want to keep, you can delete each resource individually instead of deleting the resource group.

Delete a resource group

  1. Sign in to the Azure portal, and then select Resource groups.

  2. Select the resource group to delete.

    If there are many resource groups, in Filter for any field, enter the name of the resource group you created to complete this article. In the list of search results, select the resource group.

    Screenshot that shows a list of resource groups to choose from to delete.

  3. Select Delete resource group.

  4. In the Delete a resource group pane, enter the name of your resource group to confirm, and then select Delete.

    Screenshot that shows a box that requires entering the resource name to confirm deletion.

Within a few moments, the resource group and all of its resources are deleted.

Summary

This tutorial and Get started with Azure Functions triggers in Azure Cache for Redis show how to use Azure Cache for Redis to trigger Azure function apps. They also show how to use Azure Cache for Redis as a write-behind cache with Azure SQL Database. Using Azure Cache for Redis with Azure Functions is a powerful combination that can solve many integration and performance problems.

Related content