Skip to content

An open-source GUI for managing and monitoring Kafka cluster health

Notifications You must be signed in to change notification settings

oslabs-beta/kalibrate

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Kalibrate

A free and open-source web application for managing and monitoring Kafka clusters


Table of Contents


About

Kalibrate is an all-in-one, free and open source tool to make managing and monitoring your Apache Kafka clusters a breeze. It's easy to set up and provides a friendly user experience. Simply login or create an account, then connect to your Kafka instances to get started.

With Kalibrate, you can save a Kafka cluster to your account by giving it a unique name and entering a valid URI along with SASL credentials if applicable. Add new clients to your account at any time, or remove one by selecting one of your saved clients and clicking delete.

Once connected to a Kafka cluster, you'll be greeted with an high level overview of its metadata and health. Head on over to the options within the manage tab to view and configure brokers, topics, partitions, messages, and consumer groups and members. Navigate to the monitor options to get a live and dynamically graphed view of your cluster throughput and offsets.

Kalibrate was developed with ease of use and configurability in mind. The application features a light and dark mode, user account configuration options, and an alert system that provides in-app notifications, with Slack and email integrations.

Security was a top priority in Kalibrate's developement. User account information and Kafka cluster credentials are securely stored through a combination of hashing, symmetric encryption, and careful session management.

Get started with Kalibrate by visiting our website, or read on to learn how to run Kafka in your local environment. If you're interested in learning more about our intentions, you can check out our launch article, and be sure to get the latest on Kalibrate by following us on LinkedIn and Twitter.


Running Kalibrate in a Local Environment

Initialization

Run npm install to install dependencies, followed by npx prisma migrate dev --name init to map the Prisma Data Model to your database schema, and prisma generate to generate the Prisma Client. Finally, the database can be initialized with necessary records by running node scripts/init-dv.js.

Production

Run npm run build, followed by npm start. By default, the server will listen on port 5173 and the application will be accessible on localhost:5173 in your browser.

Development

Run npm run dev. An HMR enabled dev server will listen on port 5712 and the application will be accessible on localhost:5712 in your browser. Requests to the API will be proxied to the Express server listening on port 5173 by default.

Testing

Run npm run test to run all tests. Unit and integration tests can be found in the __tests__ folder.


Running Kalibrate with Docker

Before you begin, be sure to have Docker and Docker Compose installed.

Production

Run npm install to install dependencies, followed by npm run build to output the production bundles to the dist folder.

Run npx prisma migrate dev --name init to map the Prisma Data Model to your database schema, followed by prisma generate to generate the Prisma Client and node scripts/init-dv.js to initialize the database with necessary records.

Delete the node_modules folder, and build the Docker image for the production environment with npm run docker-build-prod. You can verify that the kalibrate-prod image has been created by running docker images.

Create the production container with npm run docker-prod:run. You can verify that the container has been created by running docker ps.

By default, the Express server listens on port 5173 and the application can be viewed by navigating to localhost:5173 in your browser.

To stop the container, run npm run docker-prod:stop.

For deployment, check that the dist folder is not in your .gitignore file and that any changes to your local repository have been committed. You can then zip the application with git archive -v -o kalibrate.zip --format=zip HEAD.

Development

Build the Docker images for the development environment and PostgreSQL database with npm run docker-build-dev. This creates two images, kalibrate-dev and kalibrate-postgres, which you can verify have been created by running docker images.

Run the Docker Compose container with npm run docker-dev:up. An HMR enabled dev server will listen on port 5712 and the application will be accessible on localhost:5712 in your browser. Requests to the API will be proxied to the Express server listening on port 5173 by default. Environmental variables have been preconfigured in the Docker Compose file to connect to the local PostgreSQL database.

Stop the container with npm run docker-dev:down.

Reset Docker Environment

You can remove all Docker images and containers with npm run docker-remove-all.


Environment Variables

The following environment variables should be set in a root .env file:

  • PORT: Port for the Express server
  • DB_URI: PostgreSQL database URI, if you are not utilizing the Docker development container
  • SHADOW_DB_URI: For certain hosted databases, Prisma may require you to provide a shadow database URI
  • SALT_WORK_FACTOR: Salt work factor for bcrypt hashing
  • JWT_SECRET: JSON Web Token secret
  • ENCRYPT_KEY: Symmetric encryption key
  • SENDGRID_API_KEY: Email integrations
  • SENDGRID_EMAIL: For Email integrations
  • SG_TEMPLATE_PASSWORD: Email integrations
  • SG_TEMPLATE_ALERT: Email integrations

Demo Kafka Cluster & Services

Before you begin, be sure to have Docker and Docker Compose installed.

Running Demo Cluster

The Kafka server is configured with 1 Zookeeper and 3 brokers. To create and spin it up, run npm run demo-kafka.

Running Demo Services

Once the Kafka server is running, spin up the services by running npm run demo-services.

These client instances create the following topics:

  • Unfulfilled orders
  • Inventory
  • Payments
  • Fulfilled orders

They interact with the cluster in the following way:

  • User service: produces to unfulfilled orders, subscribed to fulfilled orders
  • Inventory Service: subscribed to unfulfilled orders, produces to inventory
  • Payment Processing Service: subscribed to unfulfilled orders, produces to payments
  • Shipping Service: subscribed to inventory & payments, produces to fulfilled orders
  • Finance Service: subscribed to payments

Reset the Cluster

Stop and remove the Kafka server by running npm run demo-kafka-reset.


Contributing

Interested in contributing? We encourage you to fork the repository and open a pull request.

Planned features for future versions of Kalibrate include:

  • Producing messages
  • Adding more cluster metrics and visualization
  • Expanding available alerts and user preferences
  • Persisting selected alerts and preferences
  • Adding OAuth authentication
  • Improving the UI
  • Optimizing application performance with improved state management and caching
  • Expanding development testing suite

Contributors


Ashlee Gafaru

Jihui Xue

Jonah Hammond

Julien Devlin

Rebecca Kwong

License

Kalibrate is MIT licensed.