The Internet of Things is a fast growing market. As more homes turn into smart homes, hundreds of devices are bought online every month. The cost and effort required for operating and maintaining platforms needed to support these products is starting to outlive the product itself. Debates on whether to improve or replace the classic client-server have come about, as companies find ways to decrease costs and enter the market faster. A movement initiated by AWS’ Function as a Service has pushed for the serverless model to replace the classic model, as a means to catch up to the increasing demand of the IoT market.
In this paper we review key differences between classical and serverless infrastructure in: