LeetCode Clone Application

This repository contains two core services:

• Server: Handles client requests and interacts with the database.
• Worker: Processes code submissions and updates the database with Accepted/Rejected status.

Both services are built using Node.js and TypeScript.

Interface App

Located in the interface folder, this app is responsible for rendering the UI and handling user interactions. It includes:

• /problems route: Displays a list of problems.
• /problems/:id route: Allows code submission for specific problems.

When code is submitted, it is sent to the worker service for processing. The worker updates the database based on the results.

Common Package

A shared package between the server and worker services, containing the Prisma schema for database connectivity.

Kubernetes Deployment with Helm

We have set up a Helm chart to deploy the application in a Kubernetes cluster. This chart includes:

• RabbitMQ as a dependency, with its own service and replicaset.
• Backend service with a corresponding replicaset.
• Worker service replicaset (which waits for RabbitMQ to be ready).

Upcoming Features

• KEDA: To enable scaling of the worker replicaset based on the size of the RabbitMQ queue.
• Prometheus: For monitoring the application.

Getting Started

Clone the repository

• Install dependencies => npm install
• Do the build => npm run build
• Run the server => npm run dev

LeetCode Clone: Deployment and Management Commands

Note: All resources are deployed in a namespace called leetcode.

Note: In this application all resources are deployed in a namespace called leetcode

Docker Commands
docker build -f apps/worker/Dockerfile -t leetcode-turborepo/worker:v2 .
docker tag leetcode/worker:latest dollysingh3192/leetcode-worker:v1
docker push dollysingh3192/leetcode-worker:v1
docker pull  dollysingh3192/leetcode-worker:v1
docker run -it --entrypoint /bin/sh dollysingh3192/leetcode-worker:v1
docker run --name my-backend-container -p 3000:3000 my-backend-app
docker run --name my-backend-container -d --rm -p 3000:3000 my-backend-app
docker ps
docker stop <id of the container>

Kubernetes Commands
kubectl get namespaces
kubectl create namespace leetcode
kubectl delete namespace leetcode
kubectl get pods
kubectl get all
kubectl get pods,svc -n leetcode

Check the logs of the backend pod
kubectl logs <backend-pod-name> --namespace leetcode

List env in pods
printenv

Port forwarding to backend service
kubectl port-forward svc/backend-service 3000:3000 --namespace leetcode

Port forwarding to rabbitmq service
kubectl port-forward svc/rabbitmq 15672:15672 -n leetcode
kubectl get service -n keda
kubectl port-forward svc/my-app-prometheus-server 9090:80 -n keda
then simply access rabbitmq management console at http://localhost:15672

HELM Commands

helm create leetcode-chart
helm dependency build
helm dependency update
helm install leetcode leetcode-chart (first time)
helm upgrade leetcode leetcode-chart (updates)
//inside chart folder
helm install leetcode . --values ./values.yaml
helm install leetcode . --values ./values-prod.yaml
helm list
helm upgrade leetcode1 leetcode-chart --namespace leetcode --debug --dry-run (shows error when trying to upgrade)
helm history leetcode
helm uninstall leetcode
helm status leetcode
helm show values prometheus-community/prometheus

helm install my-app . --values values-prod.yaml -n keda --debug --dry-run >> output.txt

Run below command from root of repository
docker build -f apps/server/Dockerfile -t your-image-name .

List image content
docker run -it --entrypoint /bin/sh IMAGE_NAME

//Inside the cluster the hostname is rabbitmq.leetcode.svc.cluster.local these settings are overridden in the helm chart using the fullnameOverride and namespaceOverride values respectively.
amqp://myuser:[email protected]:5672
                        @<service-name>.<namespace>.svc.cluster.local:<port>

minikube start
minikube stop
docker container prune
docker image rm leetcode/worker


From within the RabbitMQ pod shell, run:
rabbitmqctl list_queues
rabbitmqctl list_connections
rabbitmq-plugins list
rabbitmq-plugins enable rabbitmq_management
kubectl port-forward svc/grafana <desired_local_port>:<grafana_port>
kubectl port-forward svc/my-app-grafana 3010:80 -n keda
http://host.docker.internal:9090 in grafana to connect to the prometheus instance.
kubectl get secret --namespace keda my-app-grafana -o jsonpath="{.data.admin-password}" | base64 --decode ; echo
kubectl port-forward svc/<rabbitmq-service-name> 15672:15672 -n <namespace>
kubectl port-forward svc/rabbitmq 15672:15672 -n leetcode
http://localhost:15672

"amqp://{{ .Values.rabbitmq.auth.username }}:{{ .Values.rabbitmq.auth.password }}@{{ .Values.rabbitmq.fullnameOverride }}.{{ .Values.rabbitmq.namespaceOverride }}.svc.cluster.local:5672"
"amqp://myuser:[email protected]:5672"

kubectl patch scaledobject.keda.sh worker-scaledobject -n leetcode --type merge -p '{"metadata":{"finalizers":[]}}'
kubectl patch triggerauthentication.keda.sh keda-trigger-auth-rabbitmq-conn -n leetcode --type merge -p '{"metadata":{"finalizers":[]}}'
kubectl get triggerauthentication.keda.sh -n leetcode
kubectl logs deployment/keda-operator -n keda
kubectl get scaledobjects -n leetcode
kubectl port-forward svc/rabbitmq 15672:15672 -n leetcode
kubectl get hpa -n leetcode
kubectl get deployments -n leetcode
kubectl get pods -n leetcode
kubectl get service -n keda

echo -n "amqp://myuser:[email protected]:5672" | base64

Helpful github links
* https://github.com/dollysingh3192/docker-development-youtube-series

Learning's of leetcode system design reference:
* https://www.hellointerview.com/learn/system-design/problem-breakdowns/leetcode

FOR KEDA:  
 * https://github.com/kedacore/sample-go-rabbitmq/tree/main
 * https://maddevs.io/writeups/keda-based-rabbitmq-custom-metric-for-hpa/
 * https://medium.com/cuddle-ai/auto-scaling-microservices-with-kubernetes-event-driven-autoscaler-keda-8db6c301b18


install keda
install rabbitmq
then deployment
then scaledobject

https://keda.sh/docs/2.10/deploy/#uninstall

To install infra:
helm install my-app . --values values-prod.yaml -n keda

To Uninstall infra:
kubectl delete $(kubectl get scaledobjects.keda.sh,scaledjobs.keda.sh -A \
  -o jsonpath='{"-n "}{.items[*].metadata.namespace}{" "}{.items[*].kind}{"/"}{.items[*].metadata.name}{"\n"}')
helm uninstall my-app -n keda 


also
helm install my-app . --values values-prod.yaml -n keda
helm uninstall my-app -n keda --cascade=foreground

NOTE:
Deletion will wait until all child resources are completely removed before the parent resource is deleted.
This process may take longer compared to default deletion.
It ensures a clean and predictable deletion order based on Kubernetes' ownership hierarchy.

           +----------------------+
           | External Metrics      |
           | (RabbitMQ queue size) |
           +----------------------+
                    ↑
                    | KEDA monitors
                    ↓
           +----------------------+
           |       KEDA             |
           | Creates HPA Resource   |
           +----------------------+
                    ↑
                    | Sends scaling triggers
                    ↓
           +----------------------+
           |       HPA              |
           |  Adjusts replicas in   |
           |  Worker Deployment     |
           +----------------------+
                    ↑
                    | Modifies replica count
                    ↓
           +----------------------+
           |   Worker Deployment   |
           | Runs worker pods      |
           +----------------------+

Packet to test the keda in the rabbitmq pod up by helm

{"problemId":"670b690d3559b9c9c353bdf9","code":"function twoSum(nums, target) {const map = new Map();\n    \n    for(let i = 0; i < nums.length; i++) {\n        let n = nums[i];\n        let diff = target - n;\n        \n        if(map.has(diff)) {\n            return [map.get(diff), i];\n        } else {\n            map.set(n, i);\n        }\n        \n    }}","timestamp":"2024-11-29T09:49:38.954Z","userId":"670b69df1d799b0575866ebc"}