cloud-computing-msc-ai-examples/example-projects/counter-demo/aws-lambda-localstack/start.py

from datetime import date
import zipfile
import boto3
from botocore.exceptions import ClientError

################################################################################################
#
# Configuration Parameters
#
################################################################################################

# main change from real AWS Academy access to local emulated localstack is to change the endpoint
# other than that the same tools (boto3, aws-cli, aws-cdk etc. can be used)
endpoint_url = "http://localhost.localstack.cloud:4566"

# a bucket in S3 will be created to store the counter bucket names need to be world-wide unique ;)
# Hence we create a bucket name that contains your group number and the current year.
# The counter will be stores as key (file) "us-east-1" in the bucket (same name as our default region)
# in the bucket and expects a number in it to increase
groupNr = 22
currentYear = date.today().year
globallyUniqueS3GroupBucketName = "cloudcomp-counter-" + str(currentYear) + "-group" + str(groupNr)

# region = 'eu-central-1'
region = 'us-east-1'
functionName = 'cloudcomp-counter-lambda-demo'

# The Lambda function will run using privileges of a role, that allows the function to access/create
# resources in AWS (in this case read/write to S3). In AWS Academy you need to use the role that
# use created for your student account in the lab (see lab readme).
# see ARN for AWS Academy LabRole function here:
# https://us-east-1.console.aws.amazon.com/iamv2/home?region=us-east-1#/roles/details/LabRole?section=permissions
#
# roleArn = 'arn:aws:iam::309000625112:role/service-role/cloudcomp-counter-demo-role-6rs7pah3'
# roleArn = 'arn:aws:iam::919927306708:role/cloudcomp-s3-access'
roleArn = 'arn:aws:iam::488766701848:role/LabRole'

# For localstack you can use any role Arn and every secret and access key. Hence you can also use
# existing AWS Academy credentials to connect to localstack

################################################################################################
#
# boto3 code
#
################################################################################################


def cleanup_s3_bucket(s3_bucket):
    # Deleting objects
    for s3_object in s3_bucket.objects.all():
        s3_object.delete()
    # Deleting objects versions if S3 versioning enabled
    for s3_object_ver in s3_bucket.object_versions.all():
        s3_object_ver.delete()


client = boto3.setup_default_session(region_name=region)
s3Client = boto3.client('s3', endpoint_url=endpoint_url)
s3Resource = boto3.resource('s3', endpoint_url=endpoint_url)
lClient = boto3.client('lambda', endpoint_url=endpoint_url)
apiClient = boto3.client("apigatewayv2", endpoint_url=endpoint_url)

print("Deleting old API gateway...")
print("------------------------------------")
response = apiClient.get_apis()
for api in response["Items"]:
    if api["Name"] == functionName + '-api':
        responseDelete = apiClient.delete_api(
            ApiId=api["ApiId"]
        )

print("Deleting old function...")
print("------------------------------------")
try:
    response = lClient.delete_function(
        FunctionName=functionName,
    )
except lClient.exceptions.ResourceNotFoundException:
    print('Function not available. No need to delete it.')

print("Deleting old bucket...")
print("------------------------------------")

try:
    currentBucket = s3Resource.Bucket(globallyUniqueS3GroupBucketName)
    cleanup_s3_bucket(currentBucket)
    currentBucket.delete()
except ClientError as e:
    print(e)

print("creating S3 bucket (must be globally unique)...")
print("------------------------------------")

try:
    response = s3Client.create_bucket(Bucket=globallyUniqueS3GroupBucketName)
    response = s3Client.put_object(Bucket=globallyUniqueS3GroupBucketName, Key='us-east-1', Body=str(0))
except ClientError as e:
    print(e)

print("creating new function...")
print("------------------------------------")

zf = zipfile.ZipFile('lambda-deployment-archive.zip', 'w', zipfile.ZIP_DEFLATED)
zf.write('lambda_function.py')
zf.close()

lambdaFunctionARN = ""
with open('lambda-deployment-archive.zip', mode='rb') as file:
    zipfileContent = file.read()

    response = lClient.create_function(
        FunctionName=functionName,
        Runtime='python3.9',
        Role=roleArn,
        Code={
            'ZipFile': zipfileContent
        },
        Handler='lambda_function.lambda_handler',
        Publish=True,
        Environment={
            'Variables': {
                'bucketName': globallyUniqueS3GroupBucketName
            }
        }
    )
    lambdaFunctionARN = response['FunctionArn']

print("Lambda Function and S3 Bucket to store the counter are available.\n"
      "\n"
      "You can now run invoke-function.py to view an increment the counter.\n"
      "Try to understand how Lambda can be used to cut costs regarding cloud services and what its pros\n"
      "and cons are.\n")

# API gateways require a Pro license from localstack, as a student you can apply for
# a hobby/open source dev license, that's the same license that I used for the demo in the
# lab/lecture

# API gateway would allow getting an HTTP endpoint that we could access directly in the browser,
# that would call our function, as in the provided demo:
#
# https://348yxdily0.execute-api.eu-central-1.amazonaws.com/default/cloudcomp-counter-demo

print("creating API gateway...")
print("------------------------------------")

response = apiClient.create_api(
    Name=functionName + '-api',
    ProtocolType='HTTP',
    Target=lambdaFunctionARN
)
apiArn = response

print("API Endpoint can be reached at: http://" + apiArn["ApiEndpoint"])