A method, apparatus, and computer program product for automatically detecting changes in end-user transaction performance and availability caused by a transaction server configuration change. When a configuration change to an element in a distributed computing system is proposed, a set of instrumented synthetic transactions in the distributed computer system is initiated. Access to managed resources by the synthetic transactions is tracked to produce a list of utilized resources for the transaction. The utilized resources are then categorized according to criteria including associating the utilized resources with respective transactions. The initiating, tracking, and categorizing steps are repeated after a proposed configuration change using a selected subset of the set of transactions. The effect of the proposed configuration change is then displayed for the selected subset of the set of transactions.