Turing for "provisioned concurrency/warm" lambdas

[wsdt]

Warm lambdas are charged differently. Since we charge BOBA for Turing per call, it might be worth thinking about allowing alternative ways to charge Turing calls, since occasional calls won't make up for the real AWS costs.

AWS lambdas usually cool-down after a certain period of time.
That would be an issue for most production-grade use cases since the first Turing-Request would fail which then results in a failed transaction.

AWS supports something called "Provisioned concurrency" that basically keeps your Lambda warm.
Please note that enabling this option will result in increased fees the service, since warm lambdas are hourly charged instead of by calls.

You can read more about provisioned concurrency here.