CrypPro were running their infrastructure on the Google Cloud Platform (GCP), connecting to exchanges via the public internet. This was a workable solution for them, however, as most of their exchanges are hosted on AWS, the latency benefits of being in the same location were significant.

This, combined with cost concerns on the GCP, meant that a move to AWS was due! CrypPro approached AWS, looking for a specialised partner who could assist with an introduction to the AWS platform.

CrypPro approached AWS, looking for a specialised partner who could assist with an introduction to the AWS platform as well as offer migration consultation. They needed to prepare a baseline cloud infrastructure that the internal CrypPro IT team could expand upon, building up their services within the AWS cloud.

Our proposed migration plan provided the ability to quickly redeploy into other regions when expanding into new markets. This meant infrastructure ideation that could quickly scale, taking full advantage of AWS elasticity.

Initially a network stack was created, this was configured to allow peer-to-peer connections between regions, as opposed to centralising in one region and transiting (as is common in many other setups). This was primarily to reduce round trip latencies when traversing regional boundaries of AWS.

We implemented many of the AWS account baseline setups as well as the AWS Single Sign on as a managed service for CrypPro. This enabled streamlined identity management for all future add-ons.

We crafted detailed and thorough documentation on how to carry out common maintenance tasks on both the AWS stack and with Terraform itself. This will lead to a successful migration, away from GCP, reducing data and cloud costs significantly.

We helped the customer understand the correct approach to developing new features on multiple accounts and regions, by implementing feature checks in the IaC framework. This advantage of using IaC, combined with CICD and Git Flow, allows the customer to also easily roll back, automatically using CI, any features that may break production, and meant that the solution is able to be deployed within minutes, including new features and updates, and all changes are managed via code.