sbstjn/go-lambda-example

Example AWS Lambda function in Go and SAM

Lambda Go Example

Dependecies

$ > go get github.com/aws/aws-lambda-go/events
$ > go get github.com/aws/aws-lambda-go/lambda
$ > go get github.com/stretchr/testify/assert

Build

# Build binary
$ > make build

# Test Go Code
$ > make test

Deploy

Create .env

AWS_ACCOUNT_ID=1234567890
AWS_BUCKET_NAME=your-bucket-name-for-cloudformation-package-data
AWS_STACK_NAME=your-cloudformation-stack-name
AWS_REGION=us-west-1

Install AWS CLI

$ > brew install awscli

Command

# Create S3 Bucket
$ > make configure

# Upload data to S3 Bucket
$ > make package

# Deploy CloudFormation Stack
$ > make deploy

Usage

$ > make outputs

[
  {
    "OutputKey": "URL",
    "OutputValue": "https://random-id.execute-api.us-west-1.amazonaws.com/Prod",
    "Description": "URL for HTTPS Endpoint"
  }
]

$ > curl https://random-id.execute-api.us-west-1.amazonaws.com/Stage/people

{"data":[{"id":"d1","name":"Anton","age":31},{"id":"c2","name":"Frank","age":28},{"id":"b1","name":"Horst","age":42}]}

$ > curl https://random-id.execute-api.us-west-1.amazonaws.com/Stage/person/b1

{"data":{"id":"b1","name":"Horst","age":42}}

go-lambda-example's Issues

Why is AWS_ACCOUNT_ID needed?

I saw, that AWS_ACCOUNT_ID (declared in .env) is never used.

For what could it been used?
Or: For what did you used it?

Btw. nice example! It helped me a lot to get in touch with golang and aws lambda.

Suggestion: Upload only the dist directory

In your example (which is very nice btw.) you upload the whole repository to the S3 bucket.
I know - it is only an example. But is it really important to upload the binarys AND the source code, only to use only the binary?
Could this been for debug reasons in a productive environment?

I found with a few changes you can upload only the dist directory.
You need to delete every ocurence dist/ in the template.yml and do this to the Makefile:

package: build
		cp template.yml dist/template.yml
		cd dist && \
		aws cloudformation package \
			--template-file template.yml \
			--s3-bucket $(AWS_BUCKET_NAME) \
			--region $(AWS_REGION) \
			--output-template-file package.yml

deploy:
		cd dist && \
		aws cloudformation deploy \
			--template-file package.yml \
			--region $(AWS_REGION) \
			--capabilities CAPABILITY_IAM \
			--stack-name $(AWS_STACK_NAME)

After this, only the binarys and the template.yml is uploaded.
It was not possible to exclude the template.yml. But I think it is needed to build the cloud formation stack, so the template.yml must been uploaded, because it describes the stack.

But I found also, that the S3 bucket could be emptied, and the lambda function still works.
Could it be, that the cloud formation is build, with the handler-binarys integrated in it?
Does it not have to fetch the builds from the S3 bucket, every time an instance is build because of a lambda call?

Greetings