Retail Oder Management System - AWS Cloud Migration
Case study - Kmart Order Management System to AWS
Image source: Google Images
Kmart is Australia’s most popular on-line and instore budget retail brands. At this scale and diversity of their stock, one of the big problems they have to solve is order management, for which they utilise a spectrum of solutions. Some of them are based on legacy technologies, but still are instrumental to supporting their business.
The Order Management System (OMS) was a legacy solution, that needed to be rapidly migrated to AWS as part of the program to exit the datacentre. The cost of re-engineering the solution was prohibitively high, and the decision was made to lift-and-shift the entire solution into an environment in AWS, equivalent to that in the datacentre.
Ingest the two application servers into different availability zones in AWS via VM Import/Export
Size the new servers like-to-like with the datacentre ones
Front the application servers with an ALB
Connect the new application servers to the existing database via AWS Direct Connect
During the migration outage window switch the DNS names over to the new servers
Build an Always-On SQLServer cluster on EC2 instances (to retain the licenses), equivalent to that in the datacentre with a version upgrade
Prepare a runbook for backing up the original database to S3 and then restoring it to the new cluster in AWS
During the migration outage window:
Shut down the application servers
Backup the original database into S3
Restore the backup into the new database cluster
Change the DNS names to point to the new database cluster
Innablr’s rapid mobilisation and migration approach gave the business and technology assurance that the solution was in safe hands for a quick transition to AWS.
This enabled Kmart to proceed with their exit in rapid time while maintaining their core system services without interruption to order fulfilment.
All of this was done to aggressive timelines and ensured Christmas and New Year was a successful sales season.