I have not been able to find an answer concerning this error, so I am hoping someone can help.
Computer has LabVIEW 2013 with the NI-Industrial Communications for EtherNet/IP 1.2.1 addon with Win XP although I have verified the same setup with Win 7 x32 with the same results. I am using the EthernetIP CIP Get Attribute Single.vi and the corresponding "Set" vi.
I am using a pc with five network cards, one for the corporate network (dhcp) and four that are connected to four Rockwell PowerFlex 700S AC variable drives via CAT6 crossover cables. Each 700S drive controls a motor in an engine dyno teststand. The 700S drives have the latest available firmware installed. The four NIC/700S pairs have hard coded IP addresses with the appropriate subnet masks for proper routing. I can use the Windows "ping" command without errors. The program uses a QSM-PC architecture which I have four separate qsm loops that "talk" to the each drive, those loops constantly request three parameters from the drive, a boolean drive status array, the rpm , and the torque every 250ms using the CIP "Get" vi. That data is then sent to the main front panel via shared variables. When the operator wants to say change the rpm, a write command is inserted into the queue, and the CIP "Set" vi is used.
My problem is when I start the program, it is hit or miss on whether these loops will begin to loop through their states. If they start, the program is rock solid. When one of the loops does not start, which is most of the time, the error returned is the "-251723767 Unable to connect to the network path specified". See the attached pictures. So if I disable three of the four loops, the error never shows up. Adding another loop or more will cause the error. Slowing down the request rate does not fix the issue and it is not always the same loop that will throw the error. Each qsm loop has its own session name and its own action engine with the only common thing amongst the action engines would be the "Get" and "Set" EthernetIP VIs.
No more ideas.....Kevin
