将其他文件部署到 AWS Lambda (Python)

Question

I am attempting to upload an additional file containing an encryption secret to AWS Lambda, and am having trouble. The file is to be read from my Python script and be processed. I have tested this functionality locally, and it works just fine.

I package and upload the .zip correctly, as AWS has no problems running the script once it uploads. However, my code fails at the line that it reads my file, even though it should be in the working directory.

Is it possible to upload a file into the AWS zip deployment, and have it be read by the script?

Answer 1

I was surprised that this did not work, so I did some digging for anyone interest.

I created a simple function:

import json
import os
import random

def lambda_handler(event, context):
    selection = str(random.randint(1,5))

    with open('mydata.csv') as dogs:
        for l in dogs:
            if selection == l.split(',')[0]:
                random_dog = l.split(',')[2]

    return {
        'statusCode': 200,
        'body': json.dumps('Hello from Lambda!'),
        'cwd': os.getcwd(),
        'ls': os.listdir(),
        'random_dog': random_dog
    }

A data file:

akc_popularity,year,breed
1,2019,labrador Retriever
2,2019,German Shepherd Dog
3,2019,Golden Retriever
4,2019,French Bulldog
5,2019,Bulldog

Added them to a zip archive:

$ zip fileImport.zip importer.py
$ zip fileImport.zip mydata.csv

Created the function:

$ aws lambda create-function --function-name fileImport --zip-file fileb://fileImport.zip --handler importer.lambda_handler --runtime python3.8 --role arn:aws:iam::***************:role/Lambda_StackExchange

Triggered the function:

$ aws lambda invoke --function-name fileImport output.json
{
    "StatusCode": 200,
    "ExecutedVersion": "$LATEST"
}
$ jq . output.json
{
  "statusCode": 200,
  "body": "\"Hello from Lambda!\"",
  "cwd": "/var/task",
  "ls": [
    "importer.py",
    "mydata.csv"
  ],
  "random_dog": "Bulldog\n"
}

So, please share some code so we can dive in! FYI...I would highly recommend storing secrets using AWS Secrets Manager . It is very easy to use and keeps your hardcoded secrets out of things like version control systems. Additionally, changing your secret will not require a redeployment of your function.