LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Execution trace toolkit - any experience?

Hello,

Has anybody out there experience with the new execution trace toolkit?

In priciple, it runs, but I made some strange experiences with my tcp connections to the remote machine. Which port no's are used by the execution trace toolkit?

The second miracle is the memory required. First, I thougt to give a lot to the tool, just to get a long trace. So I gave 16 000 000 bytes and got "fatal memory error"s of the RT system instead of an upload. Then I used the default of 250 000 bytes and got about 4 ms traces, which corresponded to the start of the trace, obviously ...
Now I increased the memory by a factor of 10 and got 1.2 seconds. Just guessing, where this is.

BTW, how to handle the trace of a realistic program? I
have about 500 vi's listed and get quite slow performance at viewing the results?

Thanks for any help,

acoustics
0 Kudos
Message 1 of 2
(2,555 Views)
If all you're trying to do is get a feel for how the VIs are running with respect to each other, turn off thread tracing. Turn off detailed tracing too (off by default).

As a general rule there are many more thread events than VI events over a given interval of time, and there are even more detailed events than there are thread events on that same interval. This generality, of course, is not true when a time-critical VI is hogging the CPU because a time-critical VI can have a large hierarchy and thus generate many VI resume/suspend events all within the context of a single time-critical thread.
0 Kudos
Message 2 of 2
(2,555 Views)