Shinsegae I&C has carried out tailored architecture consulting, design and installation based on its cloud-based SI execution capabilities and AWS-based MSA architecture design capabilities.
There were three issues that need to be addressed when establishing Sellpick system. First, the system had to have high availability and flexibility, while minimizing the impact between services. Second, a DR environment with no stoppage and increased stability was needed. Third, infrastructure costs for SaaS services had to be minimized.
To address these needs, module-based design by using Kubernetes-based Docker was applied. This minimized the impact between services, to allow quick application during updates. Scale-out design that allows for automatic scaling for each service type and high availability were also made possible.
In addition, by providing AWS Aurora( a relational DB), a highly stable open source data base was realized. The overall open source development helped save costs. To ensure non-stop operation of the system and increased stability, Multi-AZ design was used to create an architecture similar to DR but more cost-efficient.
To the distribution tool developed in-house by Shinsegae I&C, an electronic approval system (approval/confirmation by a superior) function was added for an additional verification process in the Saas model services.
This minimized the impact of service changes and error points and made the configuration of infrastructure more flexible based on usage to reduce initial investment costs. (A 40% cost saving is anticipated for five years based on TCO).
The Sellpick Project used encryption in units of columns by adopting a personal data encryption solution that has been verified for many years (Cubeone).
The application firewall Wapples and IPS/IDS/Anti-Malware solution, Deep Security were also applied to address security concerns.