aws-sls-docker-deployer

Creates docker images to deploy serverless lambdas to aws linux. This is good for when you have native deps that need to be compiled to aws linux before they are used.

Basically, runs npm install / pipenv install inside the container and then runs serverless deploy

There are two "flavors" one for nodejs and one for python.

The nodejs one is set on node v18 while the python is set for v3.9

building the images

make sure you have docker installed then ./build-node.sh or ./build-python.sh

using the images

The images accept the source via mounting to a preset directory (/app). To use the image, run it:

docker run --env-file .awsenv -v `pwd`:/app sls-node-deployer --stage=dev

1. the .awsenv file should contain the environment variables:

AWS_ACCESS_KEY_ID=id here
AWS_SECRET_ACCESS_KEY=secret here

if using aws sso, then this file can be generated via:

aws configure export-credentials --profile <your profile name here> --format env-no-export > .awsenv

2. -v \`pwd\`:/app mounts the current directory to /app dir inside the container, this is where the lambda's code should be.

3. all the arguments after sls-node-deployer are passed "as is" to sls deploy command, in this example --stage=dev is passed

TODO

1. complete python implementation
2. fix code replication between node/python related files