To deploy the Lambda function we can use either the AWS Toolkit for Visual Studio or the Amazon.Lambda.Tools .NET Core global tool.
When this function is deployed an IAM role is required. The role provides AWS credentials to the Lambda function that can be used to access other AWS services. When you construct a service client from the AWS SDK for .NET without specifying credentials the SDK will locate the credentials for the assigned role.
All roles used for Lambda should have access to CloudWatch Logs to support Lambda writing logs. In this Lambda function we also need access to the Amazon Simple Email Service to send emails and read access for the DynamoDB Stream.
If you don't have an IAM role that has the required permissions then skip down to the Create Role section at the bottom of this page to create an IAM role and then return back here.
In the solution explorer right click on the ServerlessTODOList.StreamProcessor project and select Publish to AWS Lambda. This will launch the deployment wizard.
In the first page we need to specify the name of the Lambda function and identify what .NET method to call. To identify the .NET method we have to set the Assembly, Type and Method name. To Lambda this will get translated to the Handler field with a value of <assembly-name>::<type-name>::<method-name>.
Be sure the profile and region are set to the profile and region that was set at the beginning of the tutorial and is where the DynamoDB table was created.
Note the bottom checkbox to "Save settings to aws-lambda-tools-defaults.json for future deployments". With this value checked all of the settings made in the wizard will be persisted into the aws-lambda-tools-defaults.json file. This makes it so redeployments can reuse all of the previous settings as well as make it easy to switch to the Amazon.Lambda.Tools .NET Core global tool.
The IAM role that you want to use to provide AWS credentials to the Lambda function. If you created the IAM role from the Create Role section on this page then select the ServerlessTODOList.StreamProcessor role.
Configuration values can be set as environment variables for the Lambda function. For our Lambda function the FROM_EMAIL variable must be set to the address that the email will be sent from.
This controls how much memory will be allocated to Lambda the function while it is processing an event. The amount of memory you allocate also controls how much CPU power the Lambda function will get. If a Lambda function is not processing fast enough for your requirements try increasing the memory settings for the function.
How much time a Lambda function will have to process an event before Lambda will cancel the process.
If a function needs to access resources that are inside a VPC, like an RDS database, then select subnets and security groups for the VPC to allow the Lambda function to access the resources. Existing network interfaces in your VPC are mapped to a Hyperplane Elastic Network Interface (ENI) and the functions connect using it. Additional IAM permissions are required for the IAM role to create the ENI. Check out the Lambda developer guide for more info.
Dead Letter Queue (DLQ) can be either an SQS queue or SNS Topic. When a Lambda function throws an unhandled exception while processing an event Lambda will retry the event a few times. If the function still fails Lambda will send the event to the DLQ if one is assigned. This can be useful for troubleshooting and debugging failed events.
Once you push upload under the covers the tooling will execute the dotnet publish command to build
the project then create zip file of the .NET assemblies and all of its required dependencies. The
zip file is uploaded to Lambda and the function is created.
If you don't have Visual Studio or you want to automate the deployment the Amazon.Lambda.Tools .NET Core Global Tool can be used.
To ensure you have installed Amazon.Lambda.Tools execute the following command.
dotnet tool install -g Amazon.Lambda.Tools
The aws-lambda-tools-defaults.json will already have defaults for most all of the values required to create the Lambda function. The only values that need to set are the Lambda function name and IAM Role.
To deploy the function execute the command from the directory of ServerlessTODOList.StreamProcessor execute the following command.
dotnet lambda deploy-function ServerlessTODOListStreamProcessor --environment-variables FROM_EMAIL=<from email address> --function-role <iam-role-name> --profile <profile-name> --region <region-name>
Notice the FROM_EMAIL environment variable is set with the --environment-variables switch. The format of this switch is
<env-name1>=<env-value1>;<env-name2>=<env-name2>.
If you also add --persist-config-file true to the command then all of the settings set during deployment will be saved in the
aws-lambda-tools-defaults.json file. That way future invocations can be done by just using the command dotnet lambda deploy-function.
For any missing required arguments like --function-role the command line tool will prompt for the missing
required parameters.
This code can be used to create an IAM Role with the appropriate permissions required for the ServerlessTODOList.StreamProcessor Lambda function. If you already have a role then this section is not necessary.
- Getting Started
- What is a serverless application?
- Common AWS Serverless Services
- What are we going to build in this tutorial?
- TODO List AWS Services Used
- Using DynamoDB to store TODO Lists
- Handling service events with Lambda
- Getting ASP.NET Core ready for Serverless
- Deploying ASP.NET Core as a Serverless Application
- Tear Down
- Final Wrap Up
Continue on to next page: Configuring DynamoDB as an event source