efio worked on an infrastructure reimplementation project to containerize core infrastructure for Linkfire. Linkfire provides artists in the creative and music industries a smart marketing platform whose trusted partners include some of the world’s biggest names in the entertainment industry. The smart platform links over 50 million fans via their favorite social media channels to the content of artists they love via their preferred streaming apps such as YouTube, Spotify, Apple Music, and Pandora. Today, when an artist releases a new single Friday afternoon and communicates it to their fans via social media, traffic at Linkfire can reach as much as 6000 visits per second.
When major artists release new media and promote it on social media through the Linkfire suite, Linkfire must be able to manage extreme peaks of traffic globally. efio helped in the redeployment of Linkfire’s most business and performance-critical applications, containerize core infrastructure so that the platform could manage this traffic load and that platform upgrades could be rolled out quickly and safely to their cloud servers located around the globe.
The Linkfire service consists of several hundred web servers, even outside of peak hours, split among four regions globally. Application updates and improvements need to be rolled out and scaled out to all the cloud servers located globally to support the millions of users.
The containerized approach, fully automated through Teamcity and Ansible, replaced an AMI-based approach involving multiple manual steps, taking multiple hours or sometimes days to fully roll out.
The transition to the next generation infrastructure enables update rollouts to be completed in less than one hour.
We helped optimize the usage of cloud infrastructure on Amazon Web Services (AWS). Linkfire’s cloud platform is built using AWS services, specifically EC2, ALB, IAM, CloudFront, Cloudwatch, Cloudwatch logs, S3, and ECR.
The effort led to significantly shorter and safer deployment cycles and a vast performance improvement.
We paid great attention to the ability to thoroughly test and performance test the application, and make sure every bit of the setup, is described in the source code repository.
Responsibilities and accomplishments:
- Described entire infrastructure as code
- Improved performance
- Enables developers to deploy several times a day
- Containerized application with Docker
- Improved observability
efio’s consultant joined a development team of four engineers. efio contributed to the roles of DevOps engineer, Cloud architect and Site Reliability Engineer.
Tools and techniques used:
Terraform, Terragrunt, Packer, Ansible, Docker, Docker-compose, Nginx, HAProxy, Amazon Linux 2, SH, K6, Teamcity, CI/CD, AWS: EC2, ALB, IAM, Cloudwatch, Cloudwatch logs, S3, ECR