A home away from home - Soho House

Soho House & Co was founded in London in 1995 as a private member’s club for people in the creative industries. The Ned, a joint venture with Sydell Group, is one of Soho Houses exclusive establishments. The heritage building hosts nine restaurants, 252 bedrooms channeling 1920s and 1930s design, a range of men’s and women’s grooming services and a members’ club.

The Challenge:

The Soho House development team has been through a complete replatforming exercise over the past year, developing and operating applications using a microservices architecture.

They evaluated various cloud providers but AWS was the preferred choice as it gave them the production confidence they needed for their new platform. The team mainly relies on RDS outside of EC2 instances alongside ancillary services such as CloudWatch. Without RDS the team wouldn’t be able to use PostgreSQL in a reliable setup which has been a game changer for their future roadmap.

While the team has been successfully deploying various microservices on Kubernetes, they were still searching for DevOps best practices that met their needs and allowed them to run with confidence. Production stability and reliability are key in order to deliver a world class experience for their exclusive establishments.

In order to achieve production reliability and optimize deployments workflows, Soho House needed to introduce observability - a holistic approach to understanding system health.

That includes understanding your deployment successes and failures and monitoring application and infrastructure analytics in real time.

“Weave Cloud helped us build our new Kubernetes clusters with confidence. The Weaveworks team went above and beyond to make sure we have the right monitoring, but also helped us to determine a proper deployment pipeline.” said Jeroen Serpieters, Head of Engineering, Soho House

Download the case study to learn how Soho House cut operations time for Kubernetes by 75% while the engineering team increased their confidence of their Kubernetes clusters by 100%.