The following graphs show the detailed results for the summary given in Headline figures. They are typical performance profiles for Caplin Liberator. They measure how message latency and server CPU usage changes as the number of subscribing simulated end users ("clients") increases, and hence as the message update rate to clients increases. Each graph shows a different usage profile; these profiles are defined in Test details.

For the first three tests (low, moderate, and high update rates) the simulated end-users ("clients") subscribed to a random set of objects out of a pool of 1000 objects where each object was updated once every 2 seconds. The number of objects subscribed to differs between tests and therefore alters the update rate received by the clients.