TCP/IP Problems

[lmc_rick@my-deja.com]

Hi:
I have an issue that is frustrating. I am tring to use the Listen
VI that comes with LabVIEW 5.1 to establish a TCP/IP connection with a
Unix Box. I'm sure that the Unix script is correct and I'm also sure
that my LabVIEW code is good because it is very similar to the examples.

Details:
LabVIEW 5.1 running on a 333 MHz PC with WinNT 4.0 and SP4
Dedicated (AUI) Ethernet connection via Ethernet HUB to:
HP-UX 10.2 OS running on a HP744 Single Board Controller


The problem is that once I start the 'server VI' the time period is
extremely long until hookup (if it ever happens!) I can perform a
netstat (DOS COMMAND) and the results say that the PC is 'listening' on
the proper port. Then I start the Client and I monitor the netstatus of
th
e Unix box and it says that the SYN was sent, but the PC sits in a
Listening state continuously without responding. Can anyone tell me why
the long delay? Also... Sometimes, the connection on the Unix side times
out without ever connecting (Currently this is set for 2 minutes)

Are there any networking issues between Unix and the PC that I should
pay special attention to?
Another quirk is that a ping between the two boxes is not always
successful. When I physically connect each Box to the Internet I can
continuously ping anyplace on the net without any errors at all, but
when I take the boxes off of the net and physically connect the boxes
together (dedicated) there is a moderate to severe packet loss when I
ping the opposite box. Sometimes successful, sometimes not.

Any and all reponses will be greatly appreciated.
Thanks
Rick



Sent via Deja.com http://www.deja.com/
Before you buy.

[Gilbert Hangartner]

Hi Rick!

Seems that I've exactely the same problem between Mac and PC. I would be
interested if you could find any explanations or solutions.

Gilbert

>
> I have an issue that is frustrating. I am tring to use the Listen
> VI that comes with LabVIEW 5.1 to establish a TCP/IP connection with a
> Unix Box. I'm sure that the Unix script is correct and I'm also sure
> that my LabVIEW code is good because it is very similar to the examples.

[advmeas.bob.l]

I don't have a solution but I may have a work around. I had a similar problem and to solve it I moved the TCP/IP client VI from the Main GUI VI to a separate VI. I am using Labview 6.1

Originally I had a Main GUI and then a subvi that ran in a parrallel loop to monitor the TCP connection. I experienced the same problems as you but when I tested the TCP VI separately it worked. Now my code has another layer to it, a main VI that calls the TCP communication VI and main GUI VI in parallel and it seems to work.

A real solution is still being investigated.

The signals being send to my program are print queue commands from a UNIX stations.