While full support of VC running in an active/passive MSCS environment is desirable, the creation of needed cluster object and starting/stopping the VC service is easy. But stopping all VCTray.exe processes on the passive node is not as easy since VC writes all application events using an event id of zero. If the event id for service start was changed to 1 and the event id for service end was changed to 2, then these events could be easily identified between the source and event id values. It would then be easy to kill all VCTray processes when the VC service stopped so they did not raise errors on the passive node. Having distinct event ids for service start and stop permits this and other fail over related automation.