Initially i used the queuing model and executed 100 transactions. Following are the results (Simulation time was 0):
#user -1 , Test Duration 1.000(s), Session Time 0.047(s), Session Busy Time 0.047(s), # of Transaction 102 (Inculding init + end)
Then, I ran in steady mode, changing the simulation time to 1 sec. Following is the result.
#users -1, Test duration 1.000 (S), session time 0, session busy time 0, # transaction 9193
I see a major difference between two whereas both has same # of transactions, same # of users, # duration. Basically i was expecting the second to be same as the first. Any thoughts?
Well they will never(in my opinion... if i'm wrong then i'll be the first to hold my hand up) be the same as the Queue'd state waits until the end of the transaction (script execution) before starting another run. When run in 'Steady' state the number of users can be set up to start at different intervals whilst the first transaction is occuring.
Don't get confused with transactions meaning users. In SilkPerformer when you talk about transactions it is generally the number of times the script is executed/initiated.