The UDP Multicast support that comes with LabVIEW 7 has some disappointing limitations. As a listener, my application wants to join two multicast groups (say, 225.1.20.1 and 225.1.20.2) on a single UDP Port (say 49900). On the OPEN to the 2nd address group, the NI support in <<LabVIEW 7.0\vi.lib\Utility\tcp.llb\UDP Multicast Read-Only Open.vi>> will generate "Error 60: Possible reason(s): LabVIEW: The specified network address is currently in use."
The underlying reason for this is baffling. These are clearly separate network addresses, so the message is misleading. The port is the same, and this may be the "network address" currently in use. If so, it is a peculiar constraint.
On a single socket in a Unix/Linux environment, you can join multiple groups via setsockopt(IP_ADD_MEMBERSHIP). Most implementations limit on the number of groups that can be joined on one socket. (Used to be 20 for BSD derivatives). But a limit of 1?
So, here are the questions:
1) Is this a limitation of the Microsoft UDP/IP stack?
2) Has this constraint been removed in later versions (do I need an upgrade?)
3) Any suggestions on where to go from here?
Message Edited by Molly K on 02-22-2005 08:35 AM