I have a scenario that pulls an Id from a csv feed file. The csv file has 50 Ids. I currently run a simulation with 50 users all at once, so 1 id is used per user. Now I would like to run the simulation over 1 minute. The issue I come across is that I run out of Ids from the csv file before the simulation can finish. Im thinking the solution to this is to implement a custom feeder that has a pool of Ids. When a user in a simulation needs an Id it is given one from the pool. At the end of the simulation the Id is added back to the pool. If there are no Ids available, then the user waits until one is available.
Is this a good solution? Or is it just better to create more Ids than the simulation will use?