Placing an R Lambda Runtime in a Container

Consider the following runtime.R file:

parity <- function(number) {
  list(parity = if (as.integer(number) %% 2 == 0) "even" else "odd")
}

lambdr::start_lambda()

The parity function accepts a number argument and returns its parity as a named list, for example:

parity(5)
# $parity
# [1] "odd"


parity(8)
# $parity
# [1] "even"

This function can then be placed into a Docker image. An example is provided below, but the key components are:

Start from the public.ecr.aws/lambda/provided parent image, which provides the basic components necessary to serve a Lambda
Install R and dependencies, both system dependencies and R packages, including the lambdr package
Copy across runtime.R and any other necessary files
Generate a simple bootstrap which runs runtime.R with R
Set the handler as the CMD. The lambdr package interprets the handler as the name of the function to use, in this case, “parity”. The CMD can also be set (or overriden) when setting up the Lambda in AWS.

FROM public.ecr.aws/lambda/provided

ENV R_VERSION=4.0.3

RUN yum -y install wget git tar

RUN yum -y install https://dl.fedoraproject.org/pub/epel/epel-release-latest-7.noarch.rpm \
  && wget https://cdn.rstudio.com/r/centos-7/pkgs/R-${R_VERSION}-1-1.x86_64.rpm \
  && yum -y install R-${R_VERSION}-1-1.x86_64.rpm \
  && rm R-${R_VERSION}-1-1.x86_64.rpm

ENV PATH="${PATH}:/opt/R/${R_VERSION}/bin/"

# System requirements for R packages
RUN yum -y install openssl-devel

RUN Rscript -e "install.packages(c('httr', 'jsonlite', 'logger', 'remotes'), repos = 'https://packagemanager.rstudio.com/all/__linux__/centos7/latest')"
RUN Rscript -e "remotes::install_github('mdneuzerling/lambdr')"

RUN mkdir /lambda
COPY runtime.R /lambda
RUN chmod 755 -R /lambda

RUN printf '#!/bin/sh\ncd /lambda\nRscript runtime.R' > /var/runtime/bootstrap \
  && chmod +x /var/runtime/bootstrap

CMD ["parity"]

The image is built and uploaded to AWS Elastic Container Registry (ECR). First, a repository is created:

aws ecr create-repository --repository-name parity-lambda --image-scanning-configuration scanOnPush=true

This provides a URI, the resource identifier of the created repository. The image can now be pushed:

docker tag mdneuzerling/r-on-lambda:latest {URI}/parity-lambda:latest
aws ecr get-login-password | docker login --username AWS --password-stdin {URI}
docker push {URI}/parity-lambda:latest

In either the AWS console or the command line, a Lambda can be created from this image. Call the Lambda “parity” to match the function name. Tests can be executed within the console. Alternatively the Lambda can be invoked from the CLI:

aws lambda invoke --function-name parity \
  --invocation-type RequestResponse --payload '{"number": 8}' \
  /tmp/response.json --cli-binary-format raw-in-base64-out

The output is now available in the generated file:

cat /tmp/response.json

{"parity":"even"}