Tragic Tragic Software & Cloud Solutions | San Diego • Seattle Blog

Workloads at Scale: Laravel Workloads + Kubernetes Worker Pod

Reading Time:< 3 Min
Published:September 29, 2021
Last Modified:July 20, 2022

Oversight is an internal product that we built to help agencies and software teams run more efficiently. Oversight connects to all of the software systems a business already uses to analyze and visualize important data insights, which inturn, helps leaders manage their teams, projects, and business in a more effective manner. Learn more about the platform’s solutions at

As we were building out the backend we had to think through how it would manage all of the complex data and integration calls for all of the different systems used by a single business. As well as how that would scale out for thousands of users.

The backend of the application is built with the Laravel framework. We decided to leverage Laravel’s core workloads functionality which is built with Beanstalk (AWS): a simple, fast work queue; along with their advanced scheduling and caching mechanisms.

With a little configuration, Laravel workloads and scheduling allow us to send out emails and notifications on a Kubernetes worker pod which is running the beanstalkd queue system. Laravel posts to the queue on the worker pod and then consumes completed jobs from it.

By leveraging Laravel cache, we are able to limit repeat API calls to the outside systems, making the backend more performant.

This configuration of caching and scheduling will allow Oversight to scale seamlessly, keeping the API snappy and responsive while still crunching giant amounts of data for its users.

Being an internal product, Oversight allows us the freedom to push the boundaries of our favorite platforms, as well as to not be tied down by client feedback or budgets. Our team from strategy to design to engineering has been encouraged to treat this project like their own, to help us produce a product that we are all proud of. We are ecstatic with what we have built, and we encourage you to give it a demo and let us know what you think.

#Code View All