After reconnect TCP/IP messed up

[GregFreeman]

Again...back with more TCP/IP issues. When I start up my two cRIOs they connect fine, and on my host side I read the header that the cRIOs send up and properly read the following data coming in. Now, if I stop my host application, the connections are closed on both the RT side and host side, and the cRIOs continue running but go into "reconnect" states which I have verfied they do properly. I know this because I have probed the refnums and they are now zero everywhere and I have checked the state being deqeued. Now if I restart the host, the cRIOs reconnect, but when they do and I read my 9 bit header and I typecast it, I see garbage, as if it's reading the middle of a packet. This results in not type casting the rest of the string correctly, and all my data essentially being meaningless. Anyone have any idea what could be happening that is causing my packets to be read correctly, but on a reconnect go bad?

[GregFreeman · Accepted solution]

I switched it and set the chassis as the host and the PC application to connect to each one individually. The chassis now each have a listener and the host connects to each one. I changed nothing else and now everything works. Despite being in completely separate loops and connecting on separate ports, the connections must have still been stepping on each other somehow when the host was the listener. Either way, this is a fix that will work for me. It may be something to look into for future reference.

Not really a solution...but a fix, so I'll mark it anyways.