One of the best things about Flood is its ability to show you results in real time, while your test is running. This lets you see right away whether your test is running as expected and make any changes necessary to your environment.
We don't have access to your application servers, so all our logs have to do with the load that we generate: the transaction metrics from your test, as reported by the load test tool you've selected; the throughput and user load; and the health of the grid nodes.
From your side, we recommend you supplement these results with data about your application server(s), such as the resource utilization of key components. This can include CPU utilization and memory utilization at a minimum. It's good practice to set up the collection of these metrics before the test is executed, so that you can further investigate any performance bottlenecks that your flood may identify.
On Flood, there are a few places where you can view your results while a test is running:
The Project view shows you your currently running flood in comparison to other floods you've run in the same project.
The Flood view shows you metrics for the current flood in isolation, including an average for all transactions in your test.
The Transaction detail view shows you metrics for just one transaction in particular, allowing you to deep-dive on failures.
The Grid resource utilization shows you the health of the grid node(s) during a test.