How scaling works for app functions?

[david_petroni]

Hello,

I'm wondering how scaling works for our app functions. Are they spawning based on the load (so, effectively, we can have multiple run of the same function in parallel) or not?

I wasn't able to find this answer in the documentation, and while not needed for our design, it would be great to have some insights here.

Thanks!

[stefan_baumgart · Accepted solution]

You can think of our serverless platform as a cluster of independent workers where each worker takes on one workload at a time, which means that if you execute multiple requests in parallel, they will be executed isolated from all other workloads. After execution, the worker takes on a new workload. The principle is very similar to what e.g. AWS Lambda offers.

This means that you can run the same function in parallel, multiple times, but be aware that there are some yet-to-be-disclosed limits on parallel execution per tenant.