99.99% website availability via migration to AWS and infrastructure modernization

Client’s result

99.99% website reliability during sales season and special marketing campaigns

50% optimized operational efficiency

Industry |

Retail Luxury Goods and Jewelry

Company size |

201 - 500 employees

Service | Migration to AWS, infrastructure modernization, AWS infrastructure support

Location | Ukraine

About Pandora in Ukraine

Amadeo is the official distributor of Pandora in Ukraine. Pandora is one of the world’s top jewelry brands, present in more than 100 countries around the globe. The company boasts 2,500 concept stores, online stores in most countries, and 32,000 employees worldwide. It is famous for its exquisite designs and iconic charm bracelets. Pandora demonstrates its commitment to sustainability by primarily utilizing recycled silver and gold, reducing carbon emissions, and cherishing an inclusive, diverse, and fair culture.

IT-Magic provides services to Amadeo, the official distributor of Pandora in Ukraine, with 44 retail stores, over 250 employees, and hundreds of thousands of local customers.

Main challenges

Challenge 1

Stabilize the website

The website was hosted on a small cloud provider that experienced issues about 2 times a month. These could be network outages and overloads under a sudden surge in website traffic.
In addition, the infrastructure was outdated. It exhibited poor performance (3-second page load time during sales), lacked scalability and automation, utilized resources inefficiently (30% overprovisioning), and had limited flexibility and agility.
All this caused website downtime 2-3 times a month, revenue loss, and reputational risks.

Challenge 2

Modernize the system with no increase in AWS spending

The system wasn’t completely efficient due to obsolete technologies and services. It could not withstand the traffic load, especially when the company sent out dozens of thousands of emails and push notifications in mobile apps with special offers and discounts. Not only the website loaded Pandora’s servers but also accounting, financial, logistics, and other services. So the system required profound improvement.
The client wanted to modernize the infrastructure and fix all severe issues while keeping their monthly AWS bill within approximately the same limits.

Challenge 3

Keep the system maintained and prevent issues

As the client didn’t have in-house DevOps engineers, they required further support for their modernized system for the following purposes:
- Cover regular tasks
- Respond to their emergency requests quickly
- Prevent issues with performance and security
- Ensure business continuity
- Adapt the infrastructure to changes
- Facilitate innovation
The company wasn’t looking for full-time involvement in the project as there was no need to spend so much time on DevOps tasks. They wanted to receive high-quality service when required without paying a full salary and providing perks.

What we did

Solution 1

Migration to AWS

As the current website hosting was unreliable, we decided to move the system to AWS without any architecture changes using EC2. This could solve part of the problem immediately.

The client needed a quick and working solution, so we offered a lift-and-shift (rehost) migration model where a system was moved to AWS as is. This method was not perfect because it could solve only part of the problems that were available, but it was enough for the first iteration. We saved the client time and money by finishing the work within 2 weeks.

The lift-and-shift migration was done according to our step-by-step process that includes milestones, such as:

Assessment and planning
Evaluation of the cloud environment, assessing the compatibility of the current and AWS environments, defining migration goals, constraints, and timelines, and developing a migration strategy and roadmap.
Infrastructure preparation
Provisioning the necessary infrastructure resources in AWS and configuring the target cloud environment to match or exceed the requirements of the current environment.
Data migration
Identifying and migrating data from the source environment to AWS using a data migration technique called database replication that allows for avoiding system downtime.
Application migration
Packaging and deploying applications to the target cloud environment, ensuring compatibility with AWS, validating application functionality, performance, and integration with other components.
Testing and validation
Conducting thorough testing of migrated applications and workloads, performing integration testing, regression testing, and user acceptance testing, and validating data integrity, consistency, and accessibility.
Optimization and tuning
Optimizing and fine-tuning configurations and parameters, monitoring and analyzing system metrics, logs, and performance indicators.
Deployment and cutover
Planning and executing the cutover process to transition production workloads to AWS, coordinating with stakeholders and IT teams, and monitoring the deployment process closely, and addressing issues.
Post-migration validation
Validating the successful completion of migration, conducting post-migration validation tests and audits, and recording lessons learned, best practices, and recommendations for future migrations.

Establishing monitoring, alerting, and reporting mechanisms to control the health, performance, and availability of the migrated infrastructure.

This solution gave the client the following benefits:

As a result, the client received an infrastructure hosted on AWS, which increased its reliability by 200%. Outages and crashes caused by server issues stopped.

Solution 2

App Dockerization and migration to Amazon ECS

When the infrastructure was hosted on AWS, we started working on its modernization and enhancement.

The first thing that needed to be done was containerization using Docker. This tool is a market leader due to its ease of use, portability, efficiency, modularity, ecosystem, flexibility, and support for DevOps practices.

Dockerization allowed consistent behavior across different environments, isolation between applications and the underlying host system, application portability and scalability, resource efficiency, and version control.

Since the company expected a more predictable cost structure and faster time to market, we chose Amazon ECS as a container orchestration service. Its use ensured high website availability, security, flexibility, and cost-effectiveness.

One of the most important steps in infrastructure modernization was configuring auto-scaling for ECS and Aurora RDS. Once done, the website didn’t experience any downtime during the sales season or special marketing campaigns because the system adapted its resource usage based on demand without any issues and manual work from the team.

There was also an important part of the work, normally done by our DevOps engineers by default. They configured automation and CI/CD pipelines according to best practices and industry standards.

Due to these critical infrastructure components, Pandora got the following improvements:

Higher speed and efficiency in building software, integrating code changes, running tests, and deploying updates
Consistency and reliability in software development processes
Automatic execution of unit tests, integration tests, and other quality assurance checks
Faster feedback loop on code changes, test results, and deployment outcomes
Efficient scalability of software development and delivery process
Risk reduction and visibility into the entire delivery pipeline
Faster time-to-market due to frequent and incremental deployments of features and updates
Improved collaboration and transparency among the development, operations, and quality assurance teams

Solution 3

Monthly infrastructure support plan

We offered Pandora the option to choose a relevant support plan that we provide for a fixed monthly fee. The company decided to pick up our Platinum Support Plan based on their needs.

Currently, we work with the client based on this commitment.

According to it, the client receives the following services:

Regular updates and patches
Implementation of the newest best practices
24/7 availability for emergency support
15-minute response time for outages in production
Scanning for vulnerabilities
System security hardening

Ensuring fault tolerance
Chat with a team of AWS experts
Weekly calls with the team
Regular detailed infrastructure audit (costs, backups, fault tolerance, security)
Providing DevOps assistance to in-house software developers

Pandora is completely satisfied with the service as it covers all they need, and they don’t have to hire a full-time AWS and DevOps expert. If they exceed the plan’s limit, we discuss everything individually.

Platium Support Plan

Update and patches

Newest best practices

24/7
availability

15-minute response time

Scanning for vulnerabilities

Security hardening

Ensuring fault tolerance

Chat with a team

Detailed infrastructure audit

Assistance to developers

Here is one example of how our infrastructure support service works.

The client requested upgrading MySQL database 5.7 to 8.0 in the development, staging, and production environments.

When we received this request, AWS introduced one of its newest updates – blue/green deployment. We decided to explore it and use it in this case.

Blue/green deployment allowed the creation of a new (green) database, identical to the current (blue) one. These databases were constantly synchronized but the green one was inactive. We upgraded it to MySQL 8.0 and tested it. Once there were no issues, we switched databases. The green database became active and the blue one was deactivated and saved as the latest backup version.

Due to this solution, the upgrade happened with no downtime and minimal involvement of the development team in the process.

Key Results and Business Value:

#1: Complete infrastructure modernization and efficiency improvement

#2: 100% system scalability

#3: 100% adherence to best practices

#4: 2x website performance improvement

#5: Increased agility and flexibility of development by 50%

#6: Predictable budgeting of AWS spending

Features Delivered:

#1: Application reliability

#2: System scalability

#3: Infrastructure security information and event management

#4: Availability monitoring

Technology we used

Client’s feedback

“Our partnership with IT-Magic has been transformational for our business. Maintaining a stable and efficient online presence is paramount for us, and your company has played a crucial role in helping us achieve our goals. IT-Magic has exceeded our expectations. The collaboration with your team has delivered tangible results for our business.

Your expertise, dedication, and commitment to our success have made you invaluable partners in our journey towards digital excellence.”

Oleksandra Baranovska
Head of E-commerce at Amadeo