Hi stef,
the fact that the trigger is randomly delayed between 0 and (Sampling Interval) is due to the fact that the samples can be acquired only on edges of the Sampling clock and the trigger condition can occur any time between two consecutive sample clock edges.
Basically, if you are acquiring at 100 MS/s, the sampling period would be 10 ns. The trigger condition can occur any time during this 10 ns period, causing the delay you are seeing.
What you may want to try is to enable the TDC property for the trigger circuitry on the NI-5122 Scope. The TDC is a method that allows for highly accurate trigger position calculation. It has a resolution of 100ps. In the documentation for the NI-5122 (see link below) you can read that Disabling the TDC (which is disabled by default) may reduce the rearm time for the trigger but it forces the trigger to be approximated to the nearest sample clock edge (basically the behaviour you are seeing).
Here you can find all the documentation available for NI HiSpeed Digitizers
http://digital.ni.com/manuals.nsf/websearch/44EEAE2AD6F15D2086256EEC006942E3?OpenDocument&node=132100_US
As far as the CLK OUT is concerned, you should refer to the niScope Configure Clock.vi to be able to route the clock to the external PIN.
Best regards,
AlessioD
National Instruments