What do you think is going to happen at peak of Amazon needs it’s own excess capacity at the same time customers need access capacity?
Even today, much of Amazon is not on AWS.
S3 was built from the old image hosting designs, but better.
StepFunctions and SWF before it are re-designed based on the 5 internal workflow engines that existed (and still do exist).
SQS is a re-do of AMP.
DynamoDB is a re-do of Sable.
Redshift is a re-do of their internal partitioning and querying of Oracle databases but on non-Oracle databases. Their new iteration of Redshift is designed to replicate Andes + Dryad/Cradle.
The people responsible for writing and maintaining AWS Services are my coworkers…
Your comment above doesn’t seem to reflect that history, so I just wanted to make sure it was stated.
AWS is also entirely built on Amazons builder tools like Apollo and Pipelines. So it’s really unfair to say “AWS was built from the ground up”
But that doesn’t mean they just took a few extra computers they had to spare and start hosting external customers on the same metal used by Amazon retail.
The only product that I know of in recent history that came from Amazon to AWS without many changes (and without any API or automation story) is Amazon Connect.
You wouldn't say something is built "on" Github.
Your continuous deployment and continuous delivery toolset isn't really what's running a multi-tenant cloud environment.
They’re the same CPUs they just use an internal API to provision and not AWS APIs. I’m pretty sure the internal APIs also just wrap ec2 et al at this point (for backwards compatibility)
Much of the Amazon infrastructure predates AWS and was built on homegrown systems. A lot of wasn’t ported over.
If you want to see what it looks like to have an Amazon product made available to AWS and that wasn’t designed specifically for AWS, look no further than Amazon Connect. There were no publicly available APIS for years to automate anything. Every AWS native service is API first.