However, see if this helps.
First, the only time I see a task scheduled for one system execute on another is if you have duplicate GUIDS in the console. If more than one system were imaged with a computer that was not properly prepared prior to imaging, you can see this kind of behavior. Not often, so we may need more details to properly determine what is happening, like screen shots. You can go to both the target systems and look at the NS Agent and see if the GUID is the same youself or not as well.
The second issue with a task that completes but shows failures can be several things. 1) if the return code is anything but 0, by default, we show that as failed. ONLY a return code of 0 is success. 2) if the system shows failure on the main screen but shows success in a drill-down screen, then it's a known product limitation when you go beyond the scheduled time of execution (default is 30 minutes). Again, we will likely need more information for full diagnosis.