Network Communications Problems

mrhoadar

I am have real problems keeping the communication alive between NIs and touch panels and between NIs and my RMS server (about 60 systmes). When I check the NI what I usuall find is something like this with "TCP LIST"

The following TCP connections exist(ed):
1: IP=172.23.100.151:1028 Socket=30 (Active)
2: IP=172.23.100.151:1079 Socket=0 (Dead)

Although, most of the time all I find is (Dead). This happens on connections to external devices and to telnet. It even sometimes happens to network connection itself and I have to go in via RS232. I should also mention that the system that crashes the most is an NI-700 running the AMX-Extron-RMS software.

All of the systems are running v3.21.343. Is there a way to make TCP connections more robust? Would the newer firmware help anything?

My systems are on a campus network that has been going through transitions. The main one change is moving from Linksys layer one switches to HP procurve layer 3 switches and adding a Bradford network control system on top of that. We have also been having real problems with Spanning Tree broadcast storms but those storms don't seem to be at the same time as the control system lockups.

Thomas Hayes

I am having like issues across our network on campus also, Out of dire need I removed the network and patched the panel and controller together directly. The panels and controllers that I have done this too have not gone down. I was able to prove that it was indeed a networking issue. Our network department has been great and been able to fix several issues and the drop off rate has fallen( althou it still happens). If need be have your IT department contact AMX IT department and they may be able to assist you. We ended up assigning a static trunk instead of auto detect which seems to really helped.

ericmedley

When I worked at the Univesrity of Nebraska, they were in the process of upgrading the network to a Gigabit backbone and reworking the switching. While they were doing this they were also scouring the campus for rogue networks setup by the students and/or faculty within their buildings. Part of their plan was to tighten down the network and lower the possibility of nefarious activity. We found that a lot of the traffic requirements needed by AMX were not being blocked at the local level per se but were being blocked at the router level. So one could do things like ping and whatnot if your were within a switch in that area. (that side of campus or a small group of buildings) but not across campus.

What we had to do was work with one of the network admins and have them watch us as we worked with the Netlinx master and show them when things worked and when they didn't. We were eventually able to get the switches and firewalls set correctly. It took quite a lot of work, however.

TurnipTruck

I have had NI-700s lose their network connection when exposed to certain non-AMX packets on the network. I beleive it had something to do with Cisco managed switch protocol. I put the NI-700 behind a router and only allowed port 1319 traffic and have not had problems since. FYI, this was several firmwares ago, maybe the root cause has been corrected.

Hedberg

Also, look at TN 544 about UDP.