I'm using a vi that writes Date and Time to the selected Target (see attachments). (I have the problem on an RT target and on different host machines i'm using).
To reproduce :
Date changes Month = 4, Day = 4 and Year = 2005, Time = 0:18:39, run the VI = OK Month = 4, Day = 3 and Year = 2005, Time = 0:18:39, run the VI : M=4, D= 2, Y=2005, time is ok run the vi again : M=4, D=3, Y = 2005.
Time changes Month = 4, Day = 3 and Year = 2005, Time = 1:0:0, run the VI = OK Month = 4, Day = 4 and Year = 2005, Time = 1:0:0, run the VI : date is ok, time is 2:0:0 run the vi again : date is ok, time is 1:0:0
I red the posted notes about Time Zone but i think this one is not related to that.
I don't have the Set Date & Time subVIs so I can review the code. However, at a glance I noticed the error cluster is not connected between the two. Try connecting the cluster to force the operation to occur in the same order each time.
Just curious to see if this eliminates the random nature of the results.
I've actually seen this as well a while back, and even putting a wait between them didn't help (I tried a short one, I think [about a second] but it should have been enough for the network). I didn't have any use for it, so I didn't look any further. I just ran the VI twice.