Case Studies

No more junk food in Czechia!

Written by Tereza Grill | July 15, 2019

How Dáme Jídlo migrated to Google Cloud Platform and saved thousands of Czechs from hot-dogs and frozen pizza.

Problem analysis

After entering the Czech market, the business growth of Dáme Jídlo, a Czech offshoot of Delivery Hero, started to skyrocket. The number of hungry visitors was increasing, especially during lunch and dinner. Thousands of people were opening the app with the prospect of ordering a juicy burger.

The company's IT infrastructure wasn't ready for this situation. Therefore, it sometimes happened that the app couldn't handle the traffic generated by visitors. Customers were faced with an unpleasant surprise. An image of a hot-dog appeared on the screen together with an error message, which meant that the application is down. The situation had to be addressed. The management of Dame Jídlo decided to migrate from an inadequate on-premises solution to the Google Cloud Platform (GCP).

“Our servers were unavailable for long periods of time, so we were losing orders and losing money. Whenever there was an outage we had to deal with the DevOps team of our datacentre. It was consuming huge amounts of time and money especially if the outage happened during lunch or dinner.”

Petra Tříšková - Head of Development at Dáme jídlo at the time

Solution

It was important for the company to stabilize the entire application, ensure faster deployment and optimize the CI / CD pipeline. Together we decided for not only the classic lift and shift migration to GCP, but also to redesign the application architecture and move it to the Google Kubernetes Engine (GKE). This allowed for automatic scaling up and down to the required power, and in peak times, and prevented the application from crashing during peak times.

The results were:

  • 95% drop in outages (ie from hours per month to several minutes)

  • End of financial losses and greater competitiveness

  • Stable infrastructure with convenient bidirectional scaling depending on the amount of traffic

“Originally, I just wanted to make our servers more accessible and reliable. However, now we are in a stage where we can scale up and down as needed. We are much farther than I expected! ”

Petra Tříšková - Head of Development at Dáme jídlo at the time

Technical details

  • Lift and shift migration to Google Cloud Platform

  • Google Cloud Load Balancing, that allowed the web and application to run smoothly regardless of the load on the server

  • Google Cloud Storage - a safe option for storing static data such as business documents or images.

  • Google Identity and Management - technology that makes it easy to distribute or restrict access. Significantly improves data security.

  • Google Kubernetes Engine (GKE) - a complete redesign of the application architecture, resulting in a reliable distributed system that is based on cloud-native principles and runs on containers within GKE. As a result, automatic scaling up and down depending on the load can now take place, ensuring the most efficient operation with minimum latency.

  • Google Stackdriver Monitoring - provides them with insight into what is happening on servers and point out any issues that might arise before they happen. Revolgy then linked the system to Slack and SMS for 24/7 alerting, bringing a new, more transparent feel to the infrastructure.

“The introduction of new technologies has significantly improved the administrative and financial aspects of the matter. Thanks to GCP, I can see in real-time how servers are being used and how much the service costs. Switching to GCP has significantly reduced our infrastructure costs, as we no longer have to pay the data centre for DevOps or extra monitoring. ”

Petra Tříšková - Head of Development at Dáme jídlo at the time