If you are using AWS as a provider for your Service, all Resources are other AWS infrastructure resources which the AWS Lambda functions in your Service depend on, like AWS DynamoDB or AWS S3.
Using the Serverless Framework, you can define the infrastructure resources you need in serverless.yml
, and easily deploy them.
Every serverless.yml
using the aws
provider is a single AWS CloudFormation stack. This is where your AWS Lambda functions and their event configurations are defined and it's how they are deployed. When you add resources
those resources are added into your CloudFormation stack upon serverless deploy
.
Define your AWS resources in a property titled resources
. What goes in this property is raw CloudFormation template syntax, in YAML, like this:
# serverless.yml
service: usersCrud
provider: aws
functions:
resources: // CloudFormation template syntax
Resources:
usersTable:
Type: AWS::DynamoDB::Table
Properties:
TableName: usersTable
AttributeDefinitions:
- AttributeName: email
AttributeType: S
KeySchema:
- AttributeName: email
KeyType: HASH
ProvisionedThroughput:
ReadCapacityUnits: 1
WriteCapacityUnits: 1
You can overwrite/attach any kind of resource to your CloudFormation stack. You can add Resources
, Outputs
or even overwrite the Description
. You can also use Serverless Variables for sensitive data or reusable configuration in your resources templates. Please be cautious as overwriting existing parts of your CloudFormation stack might introduce unexpected behavior.
To have consistent naming in the CloudFormation Templates that get deployed we use a standard pattern:
{Function Name}{Cloud Formation Resource Type}{Resource Name}{SequentialID or Random String}
Function Name
- This is optional for Resources that should be recreated when the function name gets changed. Those resources are also called function boundCloud Formation Resource Type
- E.g., S3BucketResource Name
- An identifier for the specific resource, e.g. for an S3 Bucket the configured bucket name.SequentialID or Random String
- For a few resources we need to add an optional sequential id or random string to identify them
All resource names that are deployed by Serverless have to follow this naming scheme. The only exception (for backwards compatibility reasons) is the S3 Bucket that is used to upload artifacts so they can be deployed to your function.
We're also using the term normalizedName
or similar terms in this guide. This means dropping any characters that aren't allowed in resources names, e.g. special characters.
AWS Resource | Name Template | Example |
---|---|---|
S3::Bucket | S3Bucket{normalizedBucketName} | S3BucketMybucket |
IAM::Role | IamRoleLambdaExecution | IamRoleLambdaExecution |
Lambda::Function | {normalizedFunctionName}LambdaFunction | HelloLambdaFunction |
Lambda::Version | {normalizedFunctionName}LambdaVersion{sha256} | HelloLambdaVersionr3pgoTvv1xT4E4NiCL6JG02fl6vIyi7OS1aW0FwAI |
Logs::LogGroup | {normalizedFunctionName}LogGroup | HelloLogGroup |
Lambda::Permission |
|
|
Events::Rule |
|
|
AWS::Logs::SubscriptionFilter | {normalizedFuntionName}LogsSubscriptionFilterCloudWatchLog{SequentialID} | HelloLogsSubscriptionFilterCloudWatchLog1 |
AWS::IoT::TopicRule | {normalizedFuntionName}IotTopicRule{SequentialID} | HelloIotTopicRule1 |
ApiGateway::RestApi | ApiGatewayRestApi | ApiGatewayRestApi |
ApiGateway::Resource | ApiGatewayResource{normalizedPath} | ApiGatewayResourceUsers |
ApiGateway::Method | ApiGatewayMethod{normalizedPath}{normalizedMethod} | ApiGatewayMethodUsersGet |
ApiGateway::Authorizer | {normalizedFunctionName}ApiGatewayAuthorizer | HelloApiGatewayAuthorizer |
ApiGateway::Deployment | ApiGatewayDeployment{randomNumber} | ApiGatewayDeployment12356789 |
ApiGateway::ApiKey | ApiGatewayApiKey{SequentialID} | ApiGatewayApiKey1 |
SNS::Topic | SNSTopic{normalizedTopicName} | SNSTopicSometopic |
SNS::Subscription | {normalizedFunctionName}SnsSubscription{normalizedTopicName} | HelloSnsSubscriptionSomeTopic |
AWS::Lambda::EventSourceMapping |
|
|
Cognito::UserPool | CognitoUserPool{normalizedPoolId} | CognitoUserPoolPoolId |
You can override the specific CloudFormation resource to apply your own options. For example, if you want to set AWS::Logs::LogGroup
retention time to 30 days, override it with above table's Name Template
.
When you do overriding the basic resources, the important part is normalizedFunctionName
. There are two rules.
- Should start with uppercase character.
- The
-
will be changed toDash
,_
will be changed toUnderscore
.
Here's an example:
functions:
write-post:
handler: handler.writePost
events:
- http:
method: post
path: ${self:service}/api/posts/new
cors: true
resources:
Resources:
WriteDashPostLogGroup:
Properties:
RetentionInDays: "30"