Enable AWS Lambda functions to be written in Erlang
The erllambda library provides all functionality needed to build and
deploy fully functional AWS Lambda
functions, written entirely in Erlang.
Erlang Lambda functions implement a simple two function behavior:
-module(hello_lambda).
-behavior(erllambda).
-export([handle/2]).
-spec handle( Event :: map(), Context :: map() ) -> {ok, Body} | {error, ErrorBody}.
handle( Event, Context ) ->
erllambda:succeed( "Hello Event ~p", [Event] ).There are really two ways to get started. First you can checkout and review
the
erllambda_example
project. This is a complete working AWS Lambda written in Erlang, with the
goal of demonstrating the capabilities available as part of the framework,
and how to write Lambda functions for different purposes in AWS.
The second path is to use the
rebar3_erllambda
plugin for rebar3. This will produce a fully working erllambda project
that can be used as a starting point for any new development. This plugin
also implements additional rebar3 erllambda zip building and AWS Lambda
function packaging that simplify development.
More detailed information about developing Lambda functions using
erllambda can be found in:
The erllambda application and its supporting libraries are primarily owned by
motobob and velimir
The erllambda application is built using rebar3,
and all other dependencies are automatically pulled in when erllambda is used in other projects
rebar.config.
The following are used as part of the Erlang/Elixir package for AWS Lambda:
erllambda- this repo. Core integration point with AWS Lambda Runtime APIrebar3_erllambda-rebar3/relxplugin to build & package your application in Erlangerllambda_docker- docker images. Used for packaging your application with proper native libraries set.mix_erllambda-mix/distilleryplugin to build & package your application in Elixirerllambda_example- basic Erlang generated exampleerllambda_sam_example- sample AWS Lambda function built using AWS SAM anderllambdalibrary.erllambda_elixir_example- basic Elixir generated example
TLDR; as long as your basic Erlang environment is setup, getting started
developing erllambda should be as easy as forking the repo, and then:
git clone [email protected]:${USER}/erllambda.git
cd erllambda
git remote add upstream [email protected]:alertlogic/erllambda.git
rebar3 compile
rebar3 ct
rebar3 erllambda zip
There are two key points about running Erlang AWS Lambda functions:
Important notice: at this moment AWS Lambda native runtime has openssl 1.0.1k
while latest linux systems have 1.0.2 or even 1.1.1..
See AWS Lambda
To be able to run Erlang Lambda functions in AWS Lambda it is vital to package
your SW with Erlang built against openssl 1.0.1.
You will want to setup DockerMachine and utilize the
erllambda_docker
repo for this purpose. This will allow you to perform release builds for erllambda based
components directly from the command line.
Users can have their own Erlang dockers and have those with statically linked preferred openssl versions. However statically linking for base libraries is discouraged.
Current testing has shown that it does not make sense to run Erlang on AWS Lambda functions with less then 256MB RAM. Having 512-1024MB is optimal for most of the use cases.
Depending on a use case, erlang code loading mode can significantly affect execution performance.
To switch between interactive or embedded modes set CODE_LOADING_MODE environment variable on
AWS Lambda function creation step with a desired value:
aws --profile default --region <region> \
lambda create-function \
--function-name <your_function> \
--memory-size 1024 \
--handler <your_function_module_name> \
--zip-file fileb://_build/prod/<your_function>-0.0.0.zip \
--environment "Variables={CODE_LOADING_MODE=interactive}"
--runtime provided \
--role <role-arn>
See Erllambda Example for the step-by-step procedure to deploy your Lambda. It boils down to following steps:
- compile and build a prod profile release of your application.
rebar3 erllambda zip
- create your AWS Lambda function
aws --profile default --region <region> \
lambda create-function \
--function-name <your_function> \
--memory-size 1024 \
--handler <your_function_module_name> \
--zip-file fileb://_build/prod/<your_function>-0.0.0.zip \
--runtime provided \
--role <role-arn>
- or update your previously deployed AWS Lambda function
aws --profile default --region <region> \
lambda update-function-code /
--function-name <your_function> \
--zip-file fileb://_build/prod/<your_function>-0.0.0.zip
- and invoke it
aws --profile default --region <region> \
lambda invoke --function-name <your_function> \
--log-type Tail \
--payload '{"msg": "hello"}' \
outputfile.txt
It is however recommended to use CloudFormation based approach described in rebar3_erllambda
Contributions to this repo are always welcome. If you have an idea for improving the this or related components, please submit a github issue, or simply submit a PR directly that implements your improvement.
For complex changes, or the introduction of a major feature, it is beneficial to discuss ideas before implementing them, so that your efforts can focus on pull requests that will be accepted more easily.
As you prepare your pull request, make sure that you follow the coding conventions that exist in the files, and always make sure that all unit and common tests run. Please ensure that your contribution always adds to the coverage percentage, and does not decrease it.
If you encounter an problem, or simply have a question about using this repo, please submit a github issue.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent