---
title: " ShareThis: Kubernetes In Production "
date: 2016-02-11
slug: sharethis-kubernetes-in-production
url: /blog/2016/02/Sharethis-Kubernetes-In-Production
author: >
Juan Valencia (ShareThis)
---
ShareThis has grown tremendously since its first days as a tiny widget that allowed you to share to your favorite social services. It now serves over 4.5 million domains per month, helping publishers create a more authentic digital experience.
Fast growth came with a price. We leveraged technical debt to scale fast and to grow our products, particularly when it came to infrastructure. As our company expanded, the infrastructure costs mounted as well - both in terms of inefficient utilization and in terms of people costs. About 1 year ago, it became clear something needed to change.
### TL;DRKubernetes has been a key component for us to reduce technical debt in our infrastructure by:
* Fostering the Adoption of Docker
* Simplifying Container Management
* Onboarding Developers On Infrastructure
* Unlocking Continuous Integration and Delivery
We accomplished this by radically adopting Kubernetes and switching our DevOps team to a Cloud Platform team that worked in terms of containers and microservices. This included creating some tools to get around our own legacy debt.
### The Problem
Alas, the cloud was new and we were young. We started with a traditional data-center mindset. We managed all of our own services: MySQL, Cassandra, Aerospike, Memcache, you name it. We set up VM’s just like you would traditional servers, installed our applications on them, and managed them in Nagios or Ganglia.
Unfortunately, this way of thinking was antithetical to a cloud-centric approach. Instead of thinking in terms of services, we were thinking in terms of servers. Instead of using modern cloud approaches such as autoscaling, microservices, or even managed VM’s, we were thinking in terms of scripted setups, server deployments, and avoiding vendor lock-in.
These ways of thinking were not bad per se, they were simply inefficient. They weren’t taking advantage of the changes to the cloud that were happening very quickly. It also meant that when changes needed to take place, we were treating those changes as big slow changes to a datacenter, rather than small fast changes to the cloud.
### The Solution
#### Kubernetes As A Tool To Foster Docker Adoption
As Docker became more of a force in our industry, engineers at ShareThis also started experimenting with it to good effect. It soon became obvious that we needed to have a working container for every app in our company just so we could simplify testing in our development environment.
Some apps moved quickly into Docker because they were simple and had few dependencies. For those that had small dependencies, we were able to manage using Fig (Fig was the original name of Docker Compose). Still, many of our data pipelines or interdependent apps were too gnarly to be directly dockerized. We still wanted to do it, but Docker was not enough.
In late 2015, we were frustrated enough with our legacy infrastructure that we finally bit the bullet. We evaluated Docker’s tools, ECS, Kubernetes, and Mesosphere. It was quickly obvious that Kubernetes was in a more stable and user friendly state than its competitors for our infrastructure. As a company, we could solidify our infrastructure on Docker by simply setting the goal of having all of our infrastructure on Kubernetes.
Engineers were skeptical at first. However, once they saw applications scale effortlessly into hundreds of instances per application, they were hooked. Now, not only was there the pain points driving us forward into Docker and by extension Kubernetes, but there was genuine excitement for the technology pulling us in. This has allowed us to make an incredibly difficult migration fairly quickly. We now run Kubernetes in multiple regions on about 65 large VMs and increasing to over 100 in the next couple months. Our Kubernetes cluster currently processes 800 million requests per day with the plan to process over 2 billion requests per day in the coming months.