I am using Performer 5.1.1 and have frequent problems where the controller will appear to lose contact with the agent machines. This happens in steady-state, increasing, and all-day workloads. The controller stops updating, Abort/Kill/Stop all have no effect. However, the agent machines often continue to run, delivering load. The controller is not reporting any problems, it just stops controlling the agents.

Obviously collecting results becomes very painful as the PerfLtcAgent process and PerfRun processes must be stopped on each agent, and then all the result files manually transferred to the controller for analysis.

Anyone else experiencing this or have an idea on what the problem might be?