long story short
Fixly is an e-commerce platform that matches verified service providers with customers across various industries. Rebbix has been a part of their journey from the very beginning, taking care of UX, product development, software engineering, and infrastructure management.
Even though Fixly is a product under OLX Group, it was initially operating on a separate AWS account with its own unique DevOps environment and toolset. In alignment with the group's site reliability standards and policies, it was time for Fixly to migrate to the unified OLX tech stack. This meant moving over 40 AWS services without affecting the user experience while ensuring minimal downtime and seamless cross-team collaboration. And here’s how we tackled this task step by step:
1. inventory & prioritization
- Conducted a comprehensive audit of Fixly’s infrastructure to understand dependencies
- Prioritized components based on complexity to deal with the simplest tasks first, creating the basis for further migration
2. migration plan
- Developed detailed migration plans for each major component (databases, CI/CD, infrastructure)
- Created a comprehensive migration roadmap with all essential information — including rollback procedures — in a single document for the team's convenience
3. data transfer & synchronization
- Established networking links between old and new AWS accounts
- Set up database replication for PostgreSQL, DynamoDB, and Redis to ensure real-time data synchronization
- Transitioned data services gradually while maintaining a temporary infrastructure in the old environment
4. tooling alignment
- Rebuilt Fixly’s CI/CD pipeline using GitLab and ArgoCD
- Implemented infrastructure as code (IaC) using Terraform to enable automation and maintain consistency throughout the migration process
- Ensured compliance with OLX’s technical radar and SRE standards
5. testing & rollout
- Conducted testing in sandbox and staging environments under SRE oversight to avoid risks during the migration.
6. final cutover
- Routed all traffic and operations to the new AWS environment
key outcomes
40+
AWS services were successfully migrated
20 minutes
of downtime was all it took for the switchover
achieved complete infrastructure unification
within OLX’s centralized environment
ensured future scalability
by implementing Infrastructure as Code
client testimonial
team setup
1
Engineering Manager
3
Backend Engineers
2
Quality Engineers
3
Site Reliabiity Engineers
1
Staff Engineer
1
Head of Engineering
meet the team
tools and technologies
- AWS
- GitLab CI/CD
- ArgoCD
- Helm
- AWS cross-account tools
- DynamoDB
- Kubernetes
- Kenesis
- OpenSearch
- PostgreSQL
- DynamoDB
- NodeJs
- Go
- Terraform
lessons learned
Despite the success of this migration, we had some challenges along the way. Here’s what we learned from this experience and would recommend to other teams involved in similar projects:
- start with deep dives and structured planning
Begin with in-depth discovery sessions, Kanban planning, and well-defined steps to ensure clarity and alignment before migration starts
- establish clear leadership
A single technical lead should oversee migration processes from the very beginning to avoid miscommunication and delays
- involve all stakeholders early on
It’s important to engage all relevant teams from the start to align expectations and avoid late-stage blockers
- treat migration as a marathon, not a sprint
Migration impacts management, business, and stakeholders, so setting the right expectations early and clearly communicating about needs, challenges, and progress is crucial for alignment.
- optimize access
Use a single VPN and single role for accessing everything needed during the migration to reduce switching roles and focus.
- test your maintenance pages before the rollout
Ensure maintenance pages work reliably across platforms to prevent unexpected user impact
Need to modernize or migrate the infrastructure of your product?
.png)
