Airline Cloud migration project

Role on the project

I was the lead AWS Cloud Architect on a project to move an airline from their own data centres to the AWS Cloud. The client’s existing infrastructure was a traditional self-managed data centre environment with a significant number of bespoke applications running their online booking and reservation systems. My role was as the lead AWS architect on the project working with the client’s IT team, development teams, security & compliance teams and also the client’s CTO.

Project Successes

  • Ensured the legacy applications perform at an acceptable standard on Public Cloud infrastructure once moved from their bare metal server environments.
  • Smoothly migrated the online reservation systems to mitigate against downtime and data loss during the migration of the live domains.
  • Developed a Disaster Recovery and backup infrastructure for the client’s new Cloud-based infrastructure as their solutions in place within the tradition data centre would no longer be fit for purpose.
  • Developed a high performance Oracle database layer on the AWS Cloud that could meet or exceed the performance of their existing database platform.
  • Ensured the new Cloud infrastructure was to the standards required for client’s aviation systems compliance.
  • Educated the client on the benefits and compromises required when migrating to Cloud infrastructure from a traditional data centre model. There is a significant mindset shift required to move from traditional data centre / bare metal to the Public Cloud and this must be addressed to get stakeholder buy in from management through to IT support.

Technologies Used

  • Amazon VPC (Virtual Private Cloud) deployed across multiple availability zones to provide the resilience that would have been provided across multiple data centres.
  • AWS Elastic Compute (EC2) was used to replace the physical stacks used in the data centres for web servers, database servers and middleware.
  • CloudWatch, Autoscaling and Load Balancing (ELBs) were used to detect the health of the infrastructure and also to monitor the customer load. The solution elastically scaled and self-healed based on customer demand and in response to any server failures.
  • EBS / Provisioned IOPs were used to produce a high performance disk infrastructure for the Oracle database servers on the database tier. This solution exceeded the bare metal performance of the client’s previous on premises solution.
  • AWS CloudFormation was used to scripted deployments of the infrastructure components that could be provisioned in an automated and predictable fashion.