Transitioning from Heroku to AWS
In June of 2017 I was asked to evaluate AWS for one of our major client programs. The client was utilizing Heroku for their Cloud solution and at our urging, as well as other external influencers, leaning more towards AWS. We setup a test account under SEP’s AWS account and from there setup a mock of the product up until that point. This was accomplished first by dockerizing the services as they existed and standing them up in AWS using ECS. The code worked with NO code changes, which was huge for the client to know they could port with minimal to no impact to current development.
In addition to understanding the impact to development, various support tools were investigated such as:
- Queing – SQS
- Caching – Redis and ElastiCache
- Jenkins running in a stand-alone EC2 instance
At the same time, the client was in talks with another supplier to help them with the move to AWS. This other supplier was a global company they have worked with before. Our efforts culminated in a half day workshop with ourselves and the other supplier presenting the pros and cons for moving to AWS. Our efforts over the last month paid off in being able to showcase our knowledge in not only what AWS had to offer, but how it complemented the product we were building. The client left that workshop confident that we were the right team to help them in transitioning from Heroku to AWS.
By November of that year, Todd Trimble and myself transitioned the entire team relatively seamlessly over to AWS. In a future blog post, I will elaborate on our Jenkins integration.