Skip to content

Latest commit

 

History

History
155 lines (101 loc) · 4.67 KB

README.md

File metadata and controls

155 lines (101 loc) · 4.67 KB
Raw

FaaS for Perl with Lambda Layers

This example borrows from several existing online repositories, or tutorials. They will be referenced below.

In particular, I've forked two repositories and modified that will be used. In particular the bootstrap script has been refactored to include a context object, and use error apis.

Dependencies

  • Docker
  • AWS Account and CLI
  • AWS SAM CLI

Getting started

Clone example layers and functions

git clone https://github.com/dheffx/aws-lambda-perl5-layer.git layers
git clone https://github.com/dheffx/aws-lambda-perl5-layer-example.git functions

Build the layers

make -C layers build-docker-image PERL_VERSION=5.28.1 CONTAINER_TAG=5.28
make -C layers build CONTAINER_TAG=5.28

Note: This can take a few minutes

This will build a docker image called lambda-perl-layer-foundation, then will extract its contents as two zip files.

ls layers/*.zip
lambda-layer-perl-5.28.zip lambda-layer-perl-5.28-paws.zip

The first one is the base Perl runtime. The other is the lib for Paws.

Paws is installed because

  • To demonstrate multiple layers for a single function
  • When running functions in AWS, you probably want to access other AWS services
  • It is a large amount of files that would increase the artifact size for each function that would otherwise vendor it

The base layer image is also used for running Carton to vendor the additional libraries defined in cpanfile for each function.

See the project README.md for more.

Build the sample functions

Each sample function consists of at least a handler script, such as handler.pl. Most of them also contain a cpanfile.

The Makefile in functions/ is set up to use the image made for the layers to build their artifact. Any modules defined in cpanfile and bundled.

make -C functions build CONTAINER_TAG=5.28 FUNC_NAME=simple-yaml
make -C functions build CONTAINER_TAG=5.28 FUNC_NAME=plack

Some are just a single file and do not need the build, you can just zip them

zip -j functions/hello-world/hello-world.zip functions/hello-world/* -x *.zip
zip -j functions/s3-upload/s3-upload.zip functions/s3-upload/* -x *.zip
zip -j functions/python-system-perl/python-system-perl.zip functions/python-system-perl/* -x *.zip

The s3-upload function utilizes the 2nd layer for Paws

Create a bucket to store your function artifacts

aws s3 mb your-perl-faas-project-artifacts
export ARTIFACT_BUCKET=your-perl-faas-project-artifacts

Execute functions locally

use sam local commands to execute functions locally

https://docs.aws.amazon.com/serverless-application-model/latest/developerguide/serverless-sam-cli-command-reference.html

Workaround for known bug with SAM

aws/aws-sam-cli#947

In short, SAM is not properly extracting layers from zip files for sam local.

To get around this, unzip the layers to a temporary directory, and update template.yaml to reference them

ContentUri: layers/tmp

Without doing this, sam local will show an error of Error: Couldn't find valid bootstrap(s)

Example invocations

Direct Invoke

sam local invoke HelloWorld --event test/hello.event
sam local invoke SimpleYaml --event test/yaml.event
sam local invoke PythonSystemPerl --event test/hello.event

AWS Events

sam local generate-event s3 put | sam local invoke S3Upload

Package and deploy

sam package \
  --region $AWS_REGION \
  --template-file template.yml \
  --s3-bucket $ARTIFACT_BUCKET \
  --output-template-file ./out.yml

aws cloudformation deploy \
  --template-file ./out.yml \
  --stack-name perl-faas \
  --capabilities CAPABILITY_IAM

Test the s3 upload function

perl test/generate-files.pl outdir
aws s3 sync outdir/ s3://$UPLOAD_BUCKET/landing
aws s3 ls s3://$UPLOAD_BUCKET/processed

Reference

The two repositories that was forked for this example:

Here are a few other articles out there on this topic

I would not recommend using the bootstrap file found in these articles, but otherwise there is some good information in them.

AWS Docs