AWS cost reduction by 46% due to migration to a new infrastructure and implementation of dynamic environments
Client’s result
Cost optimization and spending reduction by 46%
100% scalable AWS services
Industry | Retail & E-commerce
Company size | 1K-5K employees
Service | AWS consulting, migration to AWS, AWS infrastructure support, AWS security services
Location | Ukraine
About the Client
Foxtrot is the largest retailer of electronics and household appliances in Ukraine. The number of its stores and sales volumes are the highest in the country. Their 124 retail stores are located in 67 cities and towns with a total area of about 154 thousand sq.m.
Since 2010, the Foxtrot retail chain has been a member of Euronics, the largest purchasing group of electronics and household appliances in 30 European countries. The company has 4600 employees, and the number of Fox Club loyalty program participants totals 14 million people.
Main challenges
Challenge 1
Improve system stability and fault tolerance
- Improve system stability and fault tolerance When Foxtrot came to us, their system was already based on AWS. However, infrastructure configurations were very basic and their website was too outdated to use the AWS cloud to the fullest.
- The client experienced issues of website slowdown, especially at times of increased traffic load that would reach 600,000 user sessions. Page load time could exceed 3 seconds 3-4 times a month, and 1-2 downtime incidents happened monthly. Its services and PostgreSQL database couldn’t scale and the system wasn’t fault-tolerant.
Challenge 2
Reduce infrastructure costs
- As the website technologies the company used at that time were out-of-date and the infrastructure wasn’t customized to the real needs of the business, Foxtrot spent money on their system inefficiently.
- The client needed an experienced cloud consultant to audit their system and revise their approach to its management and maintenance.
Challenge 3
Keep the AWS system optimized and updated
- The company required AWS infrastructure support and maintenance from qualified AWS experts who could work on their regular tasks, prevent issues, improve system security, troubleshoot in an emergency, and keep the infrastructure updated.
- They wanted to have professionals at hand when requests arose without the commitment to full-time employment.
What we did
Solution 1
AWS infrastructure from scratch for a new website
Foxtrot wanted to build a new website and we offered to set up a new AWS infrastructure from scratch with correct custom configurations that could cover all their current requirements and easily adjust to new ones.
Our team optimized database queries to reduce the load and normalize their work. We also decided to apply Amazon RDS and Redis and used Graviton-based instances as they showed better performance results for databases.
We took care of high system performance by implementing Elastic Load Balancer that distributed the load and Amazon CloudFront with S3 that improved the loading speed of static files and saved traffic.
Our team optimized database queries to reduce the load and normalize their work. We also decided to apply Amazon RDS and Redis and used Graviton-based instances as they showed better performance results for databases.
Together with the stable infrastructure, we also configured CI/CD pipelines with Jenkins and GitLab and implemented Infrastructure as Code with Terraform. These are our must-have solutions as they help development teams be more productive, improve collaboration, and ensure faster time to market.
Migrating Foxtrot to a new AWS infrastructure improved its uptime to 99.99% and enhanced fault tolerance making the system highly available.
Solution 2
Scalability implementation
Autoscaling is a must for ecommerce, especially when it comes to Black Friday, seasonal sales, and promo campaigns. So IT-Magic couldn’t omit it.
We used full autoscaling opportunities. Our experts enabled and configured autoscaling for ECS tasks.
We also applied Auto Scaling Groups to scale EC2 instances.
Once the new AWS infrastructure was ready, our team migrated the system, connected the new website, and conducted a thorough load testing to ensure the desired results.
Solution 3
Infrastructure cost optimization
As we built the new AWS infrastructure from scratch with cost optimization in mind, we applied relevant resource and cost-saving solutions from the start.
Here are the key steps that we took to achieve cost optimization:
- When creating the new infrastructure, we removed outdated resources, databases, and load balancers not used in the old one.
- We started using Savings Plans, Spot Instances, and Reserved Instances.
- The already mentioned Auto Scaling Groups that we added for scalability using spots within development environments also reduced cloud costs. Later, we configured the scheduled shutdown of development environments to save server resources at night and on weekends.
- Our team followed a good practice of using less costly Graviton-based instances for Amazon RDS and Redis databases and implemented this solution.
- Later, we configured the scheduled shutdown of development environments to save server resources at night and on weekends.
- There were about 20 T3 instances used in the old infrastructure which we replaced with 6 larger C5 instances. T3 instances consumed a lot of CPU credits and the servers started to slow down. The C5 instance type does not use CPU credits and show better performance. This solution also proved to be cheaper.
As a result, the client saved 46% of their monthly cloud spending and could redirect funds to other business needs.
Solution 4
Monthly infrastructure support plan
IT-Magic was a great match for Foxtrot’s needs in infrastructure support as we offered them our support plans for a fixed monthly payment. The client chose the Platinum Support Plan which covers up to 80 hours of our work per month that are usually used for the following tasks:
- Regular updates and patches
- Scanning for vulnerabilities
- Weekly calls with the team
- Implementation of the newest best practices
- System security hardening
- Regular detailed infrastructure audit
- 24/7 availability for emergency support
- Ensuring fault tolerance
- DevOps assistance to in-house software developers
- 15-minute response time for outages in production
- Chat with a team of AWS experts
Platium Support Plan
Update and
patches
Newest best
practices
24/7
availability
availability
15-minute
response time
Scanning for
vulnerabilities
Security
hardening
Ensuring fault
tolerance
Chat with a
team
Detailed
infrastructure
audit
Assistance to
developers
Tasks can be very different within this form of our cooperation with Foxtrot. One of the most interesting cases was about speeding up the development of their applications via the creation of isolated dynamic environments on Amazon EKS.
Every developer can start a dynamic environment that looks and functions exactly like staging or production. This environment always contains up-to-date data due to nightly database backups from production.
Dynamic environments work only on demand, which means saving costs for the client. They allow developers to work independently without interfering with one another. A developer can restart their dynamic environment or remove it from Kubernetes using a separate Jenkins job.
We also configured the shutdown of the cluster with dynamic environments at night and on weekends to maximize cost savings from this solution.
Due to this implementation, the client received great outcomes:
- Smooth onboarding of new developers into the project as they can explore it in an isolated environment without interfering with others.
- Developers’ work independence from DevOps experts in terms of creating environments.
- Cost efficiency through turning off environments when not in use.
- Improved developer productivity without conflicts and interruptions.
- Consistent data updates due to nightly backups from the production environment.
- Utilization of Kubernetes ecosystem for enhanced capabilities.
Foxtrot uses the plan opportunities actively and is satisfied with our work as their AWS infrastructure is always up and running without any severe issues.
Key Results and Business Value:
#1: Cost optimization and spending reduction by 46%
#2: 99.99% uptime as the service stability increased
#3: Improved fault tolerance
#4: 100% scalable AWS services
Features Delivered:
#1: System availability
#2: AWS services scalability
#3: Dynamic environments for developers
#4: Scheduled shutdown of development environments
Technology we used
Client’s feedback
“We at Foxtrot are incredibly impressed with the transformative impact IT-Magic has had on our business. Their expertise in AWS consulting, migration, and infrastructure support not only stabilized our systems but also significantly optimized our costs.
The monthly infrastructure support from IT-Magic has been a game-changer for us. With proactive maintenance and 24/7 emergency support, they have ensured that our AWS environment remains highly available and secure. The creation of isolated dynamic environments has particularly stood out, allowing our developers to innovate without any conflicts while maintaining cost efficiency. Overall, IT-Magic has delivered immense value, empowering us to focus on our core business needs and drive growth. We are sincerely grateful for their outstanding work and recommend them as a reliable partner.”
Yevheniia Kuchmar
BRM at Foxtrot