Once a function starts executing inside an FaaS offering, it has only a limited amount of time in which it needs to complete its execution. Most of the famous cloud service providers have a limit set on their infrastructure for how long a function inside an FaaS offering can execute. For example, if we choose the most renowned FaaS offering by AWS, the AWS Lambda, the maximum duration for which a function can execute is limited to five minutes. This limit may vary marginally on the other providers but won't be too high.

Now this makes an interesting case for us as application developers. If one of the application components that we are trying to implement as a function may take a significantly long time to execute, then we may want to re-evaluate our choice of hosting that component inside an FaaS offering.

Now we have an idea of the kind of restrictions that FaaS offerings implement. But, why do these restrictions happen and where do our functions execute inside an FaaS offering? Let's take a look.