Hi Renato,

Interesting article which raises some good points, not sure this is specifically a serverless problem though. It’s potentially easier to have a scalability problem using serverless without seeing other cracks emerging first.

Any application which has multiple endpoints and “functions” pointing to a single database will experience the same failure regardless of the tech behind it, be it a serverless architecture, kubernetes cluster running multiple services or even plain old servers and a load balancer, it all boils down to sufficient capacity planning for all components of the architecture. Having run apps at scale using all types of technology, an architecture with a single relational DB like in the example will be the bottleneck majority of the time.

Cheers,

Gavin