Case Study

Introduction

In the rapidly shifting realm of cloud services, businesses often encounter pivotal moments where they must decide whether to remain with their current cloud provider or transition to another that more closely aligns with their evolving growth strategies and technical needs. Quligy Tech recently had the privilege of assisting one of our clients in navigating this complex landscape by facilitating a migration of workloads from Amazon Web Services (AWS) to Google Cloud Platform (GCP).

Our client, who had established a robust infrastructure on GCP, found themselves in a unique position following the acquisition of a company whose operations were deeply integrated with AWS. This acquisition presented both a challenge and an opportunity: to consolidate disparate cloud environments into a cohesive, unified platform. Recognizing the strategic importance of this integration, the product leadership was keen on migrating the newly acquired AWS workloads to their existing GCP setup.

The decision to transition to GCP was driven by a desire for harmonization of the cloud infrastructure across the newly expanded organization. This move was seen as a strategic step towards enhancing operational efficiency, improving resource management, and streamlining technological processes. The journey from AWS to GCP, while complex, was an essential step in our client’s ongoing pursuit of technological excellence and business growth.

Our Process

1. Analysis: The journey began with an in-depth analysis of the existing AWS workloads. In the world of IT leadership, cloud portability is a hotly debated topic, yet it’s often a luxury that startups, in their quest for market dominance, cannot afford. Cloud providers offer powerful native services that significantly expedite product development. However, even with efforts to create abstractions for increased portability, some degree of lock-in, particularly with services like Identity and Access Management, is almost inevitable. Hence, partnering with a team proficient in both AWS and GCP was crucial for an insightful analysis.
2. Service Mapping: Our architects collaborated closely with the client’s product team to draft a comprehensive service map for the transition to GCP. While some mappings were straightforward, such as transitioning from AWS’s RDS to GCP’s Cloud SQL, and S3 to Cloud Storage, others required more nuanced discussions. The client’s AWS workloads were deployed on containers and Lambda Functions. GCP offers several options for container deployment, including Google Kubernetes Engine (GKE), Cloud Run, and Cloud Apps. After meticulous evaluation, we selected Cloud Run as our preferred container platform for its suitability to our client’s specific needs.
3. Security and Compliance: The security frameworks and integration capabilities of AWS and GCP differ markedly. We conducted a thorough review to align AWS’s user and service roles with the corresponding roles in GCP, ensuring a seamless and secure transition.
4. DevSecOps: We extended existing Terraform modules for Infrastructure as Code (IaC) implementation and developed CI/CD pipelines using Jenkins. This approach not only streamlined the development and deployment processes but also reinforced the security and compliance posture within the new cloud environment.

Implementation

• Generative AI Augmented Code Migration: Adapting to the chosen GCP services necessitated code changes. To facilitate this, we employed our proprietary tool, ReCodeAI, a generative AI-augmented code migration engine. This innovative tool significantly streamlines the migration process. Running on GCP’s Vertex AI, ReCodeAI meticulously scans the code repository, identifies AWS service integrations, and systematically transforms them into GCP service integrations. An added benefit of ReCodeAI is its capability to update existing test cases to reflect these changes. While generative AI models provide a substantial head start in migration efforts, the expertise of skilled developers remains indispensable for debugging and optimizing more intricate integrations.
• Data Migration:
  • Storage Transfer: We utilized GCP’s Storage Transfer Service for the seamless transfer of data from AWS S3 to GCP Cloud Storage, ensuring data integrity and minimizing downtime.
  • Database Migration with GCP DMS: A migration job was established using GCP’s Database Migration Service (DMS) to move data from AWS RDS to Cloud SQL. This process was supported by establishing VPN connectivity between the GCP VPC and the AWS VPC, ensuring a secure and efficient data transfer.

Key Achievements

• Time and Effort: Time and Effort Reduction: A critical factor in the success of this migration was partnering with a team that possesses deep expertise in both AWS and GCP. This expertise, combined with the innovative capabilities of our ReCodeAI tool, was instrumental in streamlining the migration process. We achieved a remarkable 30% reduction in both time and effort compared to traditional migration methods. This efficiency gains not only accelerated the migration timeline but also significantly reduced the resource allocation typically required for such endeavors.

• Enhanced Code Quality and Consistency: The use of ReCodeAI also had a noticeable impact on code quality. By automating the translation of AWS service integrations to GCP, we ensured a higher level of consistency and reliability in the migrated code. This automation reduced the potential for human error, leading to more stable and dependable cloud operations post-migration.