Entertainment Portal in the EU: Reduced Infrastructure Expenses and Significantly Increased Reliability

The client operates an entertainment portal with a core audience in Europe. The main challenges they faced were related to a significant increase in technical support costs, an increase in the number of critical software failures, and a nonlinear growth in infrastructure maintenance costs as the number of visitors increased. The client also recognized the risks of being unable to fully recover functionality after a severe infrastructure failure. An important requirement from the client was the project's completion time due to plans to attract a large volume of new audience.
Client's Main Requests:

• Conduct a comprehensive audit of the infrastructure and code together with the client's developers.
• Develop and implement a plan to reduce expenses and increase infrastructure reliability.
• Migrate key parts of the infrastructure to Ansible and Terraform.
• Complete all work within 1 month from the contract signing.

We conducted a comprehensive audit, examining the features of the CMS and software used by the client. Based on the audit results, we made the following decisions:

• Discontinue the use of a global CDN in favor of our own servers with redundancy within the EU. This allowed for a 4-5 fold reduction in user service costs.
• Update and optimize software settings. As a result, we achieved a significant increase in portal responsiveness and a reduction in the number of errors.
• Transition from manual software configuration to the infrastructure-as-code concept.
• Move data backups from the cloud to our own servers while adhering to georedundancy standards, significantly reducing costs.

Given the critical deadlines for the client, we allocated a separate group of DevOps engineers who exclusively worked on this project with the highest priority. The proposed solutions allowed for a 4-5 fold reduction in costs, and the conducted audit helped us identify several critical errors in software configuration. As a result, we increased the availability level from 94.3% to 99.2%, meeting the client's requests. The average number of tickets to the client's internal technical support decreased by 37%. Thanks to the dedicated team of engineers, we managed to complete the project in just 3 working weeks, allowing the client to timely implement plans for audience growth.