Auto-provisioning and auto-scalability are the killer features of Function-as-a-Service cloud offerings, and Azure Functions in particular.
One drawback of such dynamic provisioning is a phenomenon called “Cold Start”.
Sometimes you need to produce a result set, which would contain N rows with numbers 1...N in each row. For example, I needed to calculate some statistics per week for N weeks starting from today and going back to the past.
During my investigation of our ASP.NET application performance issue, I've found out that XmlSerializer may require a long warm-up. The first time, when it's used for a specific class (de-)serialization, can take up to 500 ms om my machine!