code_runner_backend

README

Code Runner Backend

Demo

Frontend Project Demo

NewNewRecoding.webm

Ensure you have Docker installed and run it in Docker

Note: This project must run inside Docker because it uses specific users and permissions.

Clone the repo

git clone https://github.com/ishu17077/code_runner_backend
cd code_runner_backend

API Calls

Postman collection

The Kubernetes Method

To use with more production grade microk8s instead of minikube(New New Method)

Install microk8s

sudo apt install snapd
sudo snap install microk8s --classic

Note: For Fedora use sudo dnf install snapd instead of first command

Build the image and output it using docker

1

docker build -f ./code-runner.Dockerfile . -t code-runner;
docker build -f ./warm-runner.Dockerfile . -t warm-runner;

2

docker save -o ./code-runner.tar code-runner;
docker save -o ./warm-runner.tar warm-runner;

Secret Config

Note: For new method, use this:

First create 1-code-runner-secret.yaml from 1-code-runner-secret.yaml.sample provided. (The values must be base64 encoded in the params)

cp ./1-code-runner-secret.yaml.sample ./1-code-runner-secret.yaml

Change the values of 1-code-runner-secret.yaml accordingly

Import image in microk8s

microk8s images import ./code-runner.tar
microk8s images import ./warm-runner.tar

Check if image is present

microk8s ctr images ls | grep code-runner
microk8s ctr images ls | grep code-runner

If present all good, else revert to building the image again and importing it.

Note: You must create these files from template mentioned here

Now applying kubernetes config and secrets(Optional)

microk8s kubectl apply -f ./1-code-runner-secret.yaml
microk8s kubectl apply -f ./2-code-runner.config.yaml

Deploy Everything(One shot)

microk8s kubectl apply -f .

Check the deployment

microk8s kubectl get pods

Choose any pod and describe

microk8s kubectl get code-runner-*

Note: * is any pod id you found

Enable dashboard(optional)

microk8s enable dashboard
microk8s dashboard-proxy

Note: You can use the token provided by dashboard-proxy in the web address that will be automatically opened

To call the API

Converting the programs to base64

You need to convert your entire code to base64 encoding before passing down to api

Base64 encode/decode here

Below is an example of base64 encoding

I2luY2x1ZGUgPHN0ZGlvLmg+CmludCBtYWluKCkgewogICAgaW50IHJlczsKICAgIHNjYW5mKCIlZCIsICZyZXMpOwogICAgaWYgKHJlcyAlIDIgPT0gMCkgewogICAgICAgIHByaW50ZigiWWVzXG5ZZXMiKTsKICAgIH0gZWxzZSB7CiAgICAgICAgcHJpbnRmKCJOb1xuTm8iKTsKICAgIH0KfQ==

Make sure you select encode option at the top for encoding

[!NOTE] Please make sure you use class Solution for java programs, and don't forget to add 'package main' at the top of golang code.
JSON Payload to call the api

Note: You can either use postman or curl

Request Type: POST

{host_url}/submission/test/private

{host_url}/submission/test/public

Replace {host_url} with actual url whether that be localhost or somewhere else.

Note: /submission/test/public endpoint cannot have more than 3 tests defined

{
    "problem_id": "69",
    "language": "C",
    "code": "I2luY2x1ZGUgPHN0ZGlvLmg+CmludCBtYWluKCkgewogICAgaW50IHJlczsKICAgIHNjYW5mKCIlZCIsICZyZXMpOwogICAgaWYgKHJlcyAlIDIgPT0gMCkgewogICAgICAgIHByaW50ZigiWWVzXG5ZZXMiKTsKICAgIH0gZWxzZSB7CiAgICAgICAgcHJpbnRmKCJOb1xuTm8iKTsKICAgIH0KfQ==",
    "tests": [
    {
      "problem_id":     "69",
      "is_public":      true,
      "stdin":          "12\n",
      "expected_output": "Yes\nYes",
      "test_id":        "1"
    }
  ]
}

Postman

Install Postman from here

Curl

curl -v -X POST "127.0.0.1:300080/submission/test/private" -H "Content-Type: application/json" -H "Connection: close" --data @./examples/stress_payload.json

Note: You can pipe the above command through jq to pretty print json, in simple terms add " | jq", at the end of the above code

[!NOTE] This api will be accessible through 30080 port, as it is configured to use NodePort as load balancer

For more paraller processing go to file 3-warm-runner-deployment.yaml and change 'replicas' in accordance with your computer specification.

The minikube method(Old Method)

Install Kubernetes (kubectl to create deployments preferrably through docker desktop)

Install kubernetes and kubectl & docker desktop and click on kubernetes and create kubernetes cluster

Install Containerd Runtime You can either use kind or minikube, minikube is preferrable because it has been tested.

Install Minikube Install it on your system using Minikube Tutorial. Install KIND(Kunbernetes In Docker) Install it on your system using KIND Tutorial.

[!WARNING] With KIND, you are on your own, it has not been tested

kind create cluster

Using Minikube

minikube start

Build Docker image

docker build -t code_runner . -o code_runner.tar

Load image to Minikube

minikube load ./code_runner.tar

Now apply the kubernetes conf

kubectl apply -f .

Now we need to get a port exposed from minikube

minikube service code-runner

Note: The kubernetes is set up with load balancer with different pod, so it will be accessible via 30080 port instead of 8060 with docker-desktop kubernetes, but with minikube it would be running on port provided by above command

Access minikube dashboard

minikube dashboard

[!TIP] Use the microk8s method to avoid incompabilities.

[!NOTE] The postman collection sample is attached with this program here.