Checkout Payment

Business & Non Function Requirement Assumptions

Merchants should authenticate in Checkout Identity server and get a token prior to use the Gateway
The Merchant would be of 2 types

Merchants Users - Authenticate using password
Merchant APIs - Authenticate using ClientId and ClientSecret

High scalability of the services (each component should scale horizontally acording with the demand)
Process the payment operation in an async manner
Resilient payment requests, retry in the event of network instability or internal errors from Acquiring Bank API.
Merchants can only see payments that they requested
Maximum Expiry data validates to 20 years ahead.
The Identity/Gateway APIs will be in Public Subnet
CQRS Microservices and Resources will reside in Private Subnet

This project consist of:

Checkout.Payment.Identity - Solution containing a real identity server that Authenticates the Merchant and generates a valid JWT token with claims
Checkout.Payment.Gateway - Solution containing the logic to authenticate a merchant, send and check for payment status
Checkout.Payment.Command - Solution containing the CQRS Command logic to perform command actions (Payment Request)
Checkout.Payment.Processor - Solution contining lambda ready logic to process qeueud payment requests against an AcquiringBank API
Checkout.Payment.AcquiringBankMock - Solution containing logic to mock random payment results from an AcquiringBank API
Checkout.Payment.Query - Solution containing the CQRS Query logic to perform query actions (Check payment status)
LocalStack - AWS Mock container to simulate SNS, SQS and Lambda cloud behavior.
Redis - Redis container to simulate Redis Distributed Cache persistency across Command/Query.

Architecture Logical View

Workflow Sequence Diagram

Getting Started

1. Go to the base directory of the repository and run the batch to build the docker images

	> build-containers.bat

2. Run all the containers in docker-composer

	> docker-compose up -d

(Remove -d arg if you want to see the composer containers logs in the same window)

3. Wait for localstack Ready state and Run the batch to setup localstack services (SNS -> SQS -> Lambda (Processor)

	> setup-localstack.bat

4. Kibana Logging - Access http://localhost:5601

5. Access the Payment Gateway

Using the Payment Gateway:

1. Access Gateway Swagger at URL: http://localhost:8001/swagger

Valid Merchant User Password credentials follows the given pattern

Avaliable merchants: 100

The number of merchants and its details can be updated in (Checkout.Payment.Identity UserRepository logic)

id	username	password
1-100	merchant{id}	merchant{id}password
3	merchant3	merchant3password

Valid Merchant API credentials follows the given pattern

Avaliable merchants: 100

The number of merchants and its details can be updated in (Checkout.Payment.Identity ApiKeyRepository and Client logic)

id	Api Key	Api Secret
1-100	merchant.api.{id}.key	merchant.api.{id}.secret
3	merchant.api.3.key	merchant.api.3.secret

2. Get a merchant token: (POST /v1/Authentication) using a valid credential

3. Set the JWT token in the Authorize Swagger button

4. Create a Payment in the Gateway - POST /v1/payment

5. Get a Payment Status in the Gateway - GET /v1/payment/{paymentId}

Code	Payment StatusCode Description
100	Processing
200	Succeeded
400	Rejected - Incorrect
401	Rejected Insufficient Funds
402	Rejected Card Blocked
403	Rejected Custom
500	Unexpected

Improvements Backlog

Add domain validations (FluentValidation)
Implement UnitTests to all Solutions
CodeCoverage Check during image build
End-to-End Integatrion Tests
Performance Tests using tools like Locust
Transaction-Id header Middleware to map the transaction workflow across the APIs
Use of a Common Solution Nuget package for SeedWork / Extensions / HttpClients
Add dedicated Health Check endpoints for LoadBalancing check
Add Metrics UI for each API
Setup a full CICD script to deploy in the cloud

Tools / Swagger UIs

Redis UI: http://localhost:8081
Payment Gateway http://localhost:8001/swagger
Payment Command http://localhost:8002/swagger
Payment Query http://localhost:8004/swagger
Acquiring Bank Mock http://localhost:8003/swagger

Useful commands

Manual setup for localstack (SNS + SQS + Subscription)

aws --endpoint-url=http://localhost:34566 sns create-topic --name payment-process-topic
"TopicArn": "arn:aws:sns:us-east-1:000000000000:payment-process"

aws --endpoint-url=http://localhost:34566 sqs create-queue --queue-name payment-process-queue
"QueueUrl": "http://localhost:34566/000000000000/payment-process-queue"

aws --endpoint-url=http://localhost:34566 sns subscribe --topic-arn arn:aws:sns:us-east-1:000000000000:payment-process-topic --protocol sqs --notification-endpoint arn:aws:sns:us-east-1:000000000000:payment-process-queue

aws sqs get-queue-attributes --queue-url http://localhost:34566/000000000000/payment-process-queue --attribute-name QueueArn --endpoint-url http://localhost:34566

Generating .NET Core Processor Lambda Content to localstack

dotnet publish Checkout.Payment.Processor\Checkout.Payment.Processor.sln -c Release -o ZipLambda
powershell Compress-Archive ZipLambda\* -DestinationPath CheckoutPaymentProcessor.zip -Force

Create Processor lambda

aws --endpoint-url=http://localhost:34566 lambda create-function --function-name=checkout-payment-processor --runtime=dotnetcore3.1 --handler=Checkout.Payment.Processor::Checkout.Payment.Processor.Function::FunctionHandler --zip-file fileb://CheckoutPaymentProcessor.zip --role=r1

Update Processor lambda code

aws --endpoint-url=http://localhost:34566 lambda update-function-code --function-name=checkout-payment-processor --zip-file fileb://CheckoutPaymentProcessor.zip

Subscribing Processor lambda to SQS

aws --endpoint-url=http://localhost:34566 lambda create-event-source-mapping --event-source-arn arn:aws:sqs:us-east-1:000000000000:payment-process-queue --function-name checkout-payment-processor

Debug/Troubleshoot Lambda Logs

aws --endpoint-url=http://localhost:34566 logs describe-log-streams --log-group-name "/aws/lambda/checkout-payment-processor"
aws --endpoint-url=http://localhost:34566 logs get-log-events --log-group-name "/aws/lambda/checkout-payment-processor" --log-stream-name "yyyy/MM/dd/[LATEST]12345678"

Troubleshooting

ElasticSearch / Kibana does not start

For windows users a pre-config might be needed in wsl to allow elasticsearch container to start. Execute the 2 steps below if your elasticsearch container is not starting

Open command prompt to access access wsl bash

	C:\> wsl -d docker-desktop

Increase VM max map count to 262144:

	# sysctl -w vm.max_map_count=262144

Restart the elasticsearch container

	docker-compose up elasticsearch

RogerBarreto / checkout.payment