How to run a AWS lambda via SAM local that writes to an S3 bucket?

Question

I'm looking at running some of our AWS lambdas locally via SAM, including a one that writes to an S3 bucket. Is there a way of getting S3 to run locally, or talk to an S3 bucket in the cloud and write to that while running the lamda locally?

I did several internet searches, but was unable to work out how to do it. I've got a lambda running talking to a local dynamodb instance but not s3. — jamesdwalker, Aug 07 '19 at 10:41
@jamesdwalker which error did you get when trying to send a file into your s3? local s3 is not supported yet but you can connect into your s3 cloud service — guijob, Aug 07 '19 at 15:31

score 5 · Accepted Answer · answered May 17 '20 at 23:08

Yes, the way I achieved this (after much difficulty, the documentation is scattered all over the place and focused on specific use cases that weren't mine) was with localstack.

You need docker running, then:

pip install localstack

Then:

localstack start

Some documentation I've found suggested that you use http://localhost:4566 and indeed pointing a browser there shows something exists, but I got no luck pointing boto3 or another AWS client there, getting very cryptic error messages.

The real port is 4572, try this out:

aws --endpoint-url=http://localhost:4572 s3api create-bucket --bucket mybucket --region us-west-1

And now if you have a lambda function like this:

import boto3


def lambda_handler(event, context):
    s3 = boto3.client('s3' ,
        endpoint_url='http://localhost:4572',
        use_ssl=False)
    s3.create_bucket(Bucket="test-bucket")
    #rest of your function here

The final component (might depend on your Docker setup), is you may need to invoke it like this:

sam local invoke  --docker-network host

How to run a AWS lambda via SAM local that writes to an S3 bucket?

1 Answers1