When load testing a database cluster it is important to determine to what degree your current system can scale. The current "real" mode is useful, but I would love to see the ability to replay real traffic with a replay rate multiplier.
If I expect for example that my production traffic is going to triple over the next 12 months, I would like to run current production load first at 3 times the current rate to confirm we can handle. I would then scale that up to the point where performance is no longer acceptable to understand where the breaking point is.
When load testing a database cluster it is important to determine to what degree your current system can scale. The current "real" mode is useful, but I would love to see the ability to replay real traffic with a replay rate multiplier.
If I expect for example that my production traffic is going to triple over the next 12 months, I would like to run current production load first at 3 times the current rate to confirm we can handle. I would then scale that up to the point where performance is no longer acceptable to understand where the breaking point is.