About my PDF-Editor Project

About my PDF-Editor Project

A Web-based PDF-Editor

Dipankar Das's photo
Dipankar Das
·Sep 1, 2022·

6 min read

Play this article

Table of contents


It all started with me thinking about a PDF platform that people can use instead of using licensed software. Came up with an idea of web-based which can be used on any device be it laptop or mobile, it is lightweight by default.






For the first few versions, it was a monolithic application, later converted to a microservice architecture


  • Helps in understanding JavaScript, Golang codebase
  • A platform to implement DevOps methodology
  • To make it useful to the end-user
  • To use unit test of Golang and JavaScript

Tech stack

  • Go
  • JavaScript
  • Docker

All the files mentioned are in github.com/dipankardas011/PDF-Editor

Followed the DevOps roadmap

Cloud Native Tools

Before going any further you need to know about the different components and how they communicate.


Folder structure image.png


First lets start with Dockerfiles, here we used multiple stages for different use cases. To build and run multiple containers at once, use the Docker-compose

Screenshot from 2022-08-02 23-36-15.png from src/backend/merger/Dockerfile In the above figure

prod-stage1 -> build the executable for prod stage to consume

prod -> alpine image(reduced functionality and size) copy executable and templates directory to production image so that we can run as a production app

dev -> for development use where we mount the ${PWD}:/go/src/ and then work on the project without having to rely on the local computer go installation

test -> for testing the code using the docker image

Here in the prod stage we don't require the go binary only need the template and exec file for running the app and thereby reducing the final built image

For Docker workshop


As there were multiple containers I needed something to orchestrate them, docker-compose should work but I wanted something robust, so learned Kubernetes.

Created a deployment for backend and frontend, also learned how to attach volumes to the deployment for the backend container, added init-containers to load the data into the volume.

deploy/backend -> backend YAML files

deploy/frontend -> frontend YAML files

deploy/monitoring -> tracing, monitoring, visualization YAML files

Created the ingress rules (Used Nginx ingress) to connect to my app locally without relying on port-forward & Nodeport

Connected the persistence volume and persistence volume claim for the backend.

Used minikube as a production cluster (pdf-editor) (local system) as its free and minikube clusters can be stopped and started keeping Kubernetes data persistent which is not the case with kind.

minikube start --memory='3Gi' --nodes=2 -p pdf-editor


In the above image it has Argo rollouts deployment so you see canary

Kustomization for controlled deployment to cluster


Helm charts: Fun fact are not needed but created and deployed to the artifact Hub.

For basic deployment with no rollouts I used deploy/cluster/backend and deploy/cluster/frontend there is also deploy/cluster/monitoring for monitoring and tracing deployments


Want to get started with Kubernetes? Check out this workshop.

Argo CD

Next I learned about Argo projects and implemented a simple ArgoCD. Also used advance concepts such as Argo Rollouts to have Blue-Green rollouts and Canary rollouts also configured & used Ingress

Added Horizontal pod autoscaler image.png

using labels we create an event that when target CPU Utilization >= 80 then add more pods max of 4 or 6 in case of backend path of all the configs deploy/canary

Canary rollouts for Frontend and Blue-green for backend components




AWS with Terraform

Used Terraform Cloud to provision the infrastructure to deploy the latest version of the app to AWS EC2 instance for Staging



EC2 instance dashboard

Screenshot from 2022-08-04 18-53-47.png


Azure Web apps

Used Docker-compose for Production

It is a manual deployment as docker-compose Linux app web is not available in IAC

Screenshot from 2022-08-04 18-41-21.png




For the validation checks of Helm, Kustomize, YAML files I used Datree.


Example of one of them



Used Synk for vulnerability report of the entire GitHub Repo and Docker repo as well



Used for container vulnerability scans



Used to have policies imposed on the K8s cluster, like no use of Nodeport for security reasons

Used for policy enforcement, like no pod should have escalated privileges





It's a CI/CD tool. It can work with any software and can also be integrated with Kubernetes Implemented a simple CI Jenkinsfile pipeline

Docker runs the Jenkins server container and then attached my host computer as an agent so that we can use all the tools available


It listens to the GitHub main branch image.png




Ansible in Jenkins

Used Ansible Playbooks to configure and update the staging environment to latest alpha and beta release



Config are in deploy/IAC/ansible-terraform

Here is learning material


Used creating CI/CD Pipeline with the help of K8s

Created tasks and pipeline for trying out the tool

present in deploy/tekton-ci

created task (basic execution unit in tekton) then used them in the pipeline

Signing with Cosign

Used to sign the Docker Images



All the files are in deploy/cluster/monitoring

Screenshot from 2022-08-04 19-09-32.png


Used to get telemetry from the application of interest. With custom scrape metrics config for application

Screenshot from 2022-08-04 19-18-16.png





it is a visualization tool for all kinds of data sources be it databases, time series, logs, traces.

Created custom dashboard to count the number of 200, 500 HTTP status codes received




Used OpenTelemetry and Jaeger for tracing backend calls


Screenshot from 2022-08-04 19-09-32.png



Used the platform for automated Unit, Integration testing

Screenshot from 2022-08-06 23-07-47.png



  • Used the Terraform Cloud for the Terraform workflows
  • For the continuous testing used the standard Jest and GO testing libraries and added them to the Github actions with trigger for the Pull requests and Push to main branch
  • To use multiple Automation tools added CircleCI for test cases
  • added PR and Issue templates

Finally, these tools really helped me out

  • VSCode for all the coding
  • Lens K8s IDE Amazing tool for managing your K8s using a UI image.png

  • Monokle by Kubeshop

image.png Helped with adding correct labels for pods and their corresponding services

Currently working on -

  • more than 2 PDF merge
  • Rotate PDFs and more

Further Links: SRS Github Repo

Finally thanks to Kubesimplify for allowing me to showcase my project🙏 Hope this blog is helpful 🥳

Follow Kubesimplify on Hashnode, Twitter and Linkedin. Join our Discord server to learn with us.

Did you find this article valuable?

Support Kubesimplify by becoming a sponsor. Any amount is appreciated!

See recent sponsors Learn more about Hashnode Sponsors
Share this