Why are my TPS values the same when running the TPC-C tests on multiple servers even though the BPS values are different?
Even though the mix of transactions is different for the each servers, the number of transactions executed happens to be the same. When you add up the total number of transactions for each tests, the transaction numbers should add up to almost the same number. You can verify this by looking at the "Mix" node in each results report for each tests. That is why the TPS values are the same. They are not stressing the server so the transactions are handled faster then they are sent, limiting the TPS to the number of transactions sent not the speed of the machine.
