RT Target Network Shared Variables Error 1950679035

[Gwille]

Greetings,

I am attempting to communicate between a PC with windows XP, Firewall service disabled, LabVIEW 8.6.1, and another PC running LabVIEW RT 8.6.1. 

I intermittantly get error 1950679035 when attempting to write to the network variable from the target machine. no matter what settings I put for the network variable, or which machine I deploy the network variables on.

I have tried lots of different settings all with little success. and searched google and ni dz for any references to this problem.

I have observed this problem since the first day I installed RT, the very first VI I made just to test RT.

VIs run on the target machine run flawlessly without the use of network variables, so it is not a communication error. TCP IP connections work flawlessly too.

If I run the VI on another windows machine attached to the same network, it runs flawlessly.

I will attach an example VI which causes the error. Inside the attached project, Deploy NSV.lvlib, then start Host_ReadNSV on the host PC, and then start Target_WriteNSV on the target RT PC. after about a minute the target machine will start throwing errors and the host machine will start revealing errors in the transmitted signal.

How do you add an attachment?

Thanks for your help!

[Gwille]

I hope all of the dependencies are included

[Gwille]

--

Update

I have 5 machines running RT, using a VI very much like the exmple I posted, I tested each one of them. It seems that on one of them, the problem is not present, but on the other four it throws errors every few seconds. It is especially apparent if I remove the flushSharedVariabledata.vi, in which case the network variable is updated only about once per second. If I look at the distrubuted system manager the status of the variable shows "Disconnected" most of the time. This is all true no matter if the shared variable is deployed from the development machine or the RT target.

This one machine just magically is immune to the problem... its strange because they are all identical and I made sure that they all have the same software. I just looked in MAX: Ethernet drivers are the same, Labview RT 8.6.1, Network Variable engin... everything. 

 --

Several hours later

I booted one of the machines from a thumbstick and loaded safemode. I erased all files on the hard drive and replaced them with the files that were on the working machine.

Lo and behold it works correctly concerning network shared variables! however it reports that SMP (multicore support) is not enabled, even though it had been enabled on the orginal working machine and MAX still reports that it is in fact installed under software. sooo wtf?

OK so then using a third machine I simply disabled multicore support and it suddenly works correctly with respect to network variables.

What is it about multicore support that breaks shared variables?

how come it works properly with SMP + shared vars on that one orginial machine? 

I would prefer to run WITH SMP on all the machines... I want it all. Why is there always a problem? Does anyone have any idea whats going on? Is there any reason to believe this fix will eventually fail?

[Peñata]

Hi Gwille,

What kind of RT targets are you using? When you did your testing, did you have all of your RT targets in the same project? Do you see this same behavior if you build your VI into a real-time startup executable.

You may have already seen this KB, but I wanted to point it out anyway. Do any of the things mentioned apply to your system?

Aaron P

National Instruments

Applications Engineer

http://www.ni.com/support