This is a work related question, I primarily do Mechanical/instrumentation/hardware side of things but my data guys inherited some problems with our DAQ they can’t seem to resolve. Hardware is National Instruments, Software is NI Lab View.

Problem is we have 6 DAQ modules that are synchronized (supposedly) by a local time server that is synced via GPS.

The DAQ modules are drifting in time and these guys can’t seem to figure out the cause. Anyone have any experience with these systems?

I've used it quite a bit, mostly for frequency-domain test and instrumentation.  Haven't done anything complex enough to need the time server, though.  
Are all the DAQ units physically separated such that you need to use GPS to "remotely" sync them, or do you just need to clock stability of GPS?  If neither applies, you can always have one unit output a master clock signal that the others can synchronize against, although it won't be as low-jitter as a GPS clock.

To be honest, and this is an opinion I've heard from a whole host of other engineers- the best solution to problems involving Labview is to drag the application to the trash can.  
I understand that's rarely an option though, especially with expensive legacy hardware.

I agree completely with trashing LabView. IMO, the whole DAQ system should be scrapped, unfortunately that is not an option for a number of reasons. I order for the post-processed data to be useful, time stamps of the streams need to be kept within one second.

I don’t understand why, but we are syncing using 1588 time protocol, and we have to add offsets to account for differences in UTC, TAI and GPS times. Additionally at least on module is using its own system time that is being updated from another synced network. It’s a fucking mess of shit hardware and software coding that has been cobbled together over the years.