When switching from throughput probing (single ticket pool) to execution control with prioritization (two ticket pools), the total number of tickets is not preserved. We ignore the tickets that throughput probing had already reached in the optimal state and instead use the values from the server parameters (e.g., storageEngineConcurrentWriteTransactions, storageEngineConcurrentWriteLowPriorityTransactions, etc.).

We should evaluate whether we want to preserve the total number of tickets when transitioning from throughput probing to prioritization.

One example could be:

Throughput probing: (20 tickets)
-> 
Prioritization: (20 tickets in the normal pool, 0 tickets in the low-priority pool, heuristic disabled)