Compactlogix Wire Frame to Factorytalk SE

Hello all, I am stumped. Here it goes;

I have a 1769 L33ER (PLCA) connected to a Stratix switch A with the latest firmware 15.2(4) non ******. The switch connects to a panelview plus and 4 1794-AENTs. This switch is feed from another switch(B) in the same control panel. Switch B is a Stratix 5700 with the same firmware. This switch talks to a Compactlogix PLC B as well. PLC A is wire framing but for the wire framing periods I can ping the PLC processor from the Server and none of the local Ethernet I/O drops out in addition the local panelview plus remains polling data. But the clients wire frame.

Things I have tried. Upgraded all switches to Stratix 5700.
Rebooted both redundant Virtual Server Computers. These contain our data servers and HMI server etc....

I direct reference a tag on this screen without using HMI tags and it wire framed.

I used Factory Talk Live Data test client and the data drops.

I exported the tag database and re-imported. No help.

I re created the shortcuts in Rslinx Enterprise.

I have the ports set to autonegotiate and I have been told I should try hard setting them to 100 Full.

I have played with the overhead time slice.

From the switch connected to the processor having issues I can always ping the server even when wire framing.

Previously we had unmanaged N-tron switches installed so they must have been set to autonegotiate.

I wonder if the processor could be bad. Can anyone tell me what messaging Rockwell uses for panelviews and Aent modules and if that is different for the messages to Factorytalk?

Could only part of the PLC internal Nic card be failling?

Any else I can try. Rockwell recommended Smart ports on the Stratix Switches but this is the only processor in the plant that is having this issue.

By wire framing I'm guessing you mean the connection between PLC A and FTV SE drops out?

How loaded is the PLC, both the CPU and communication connections? Does it have a lot of MSG instructions? Can you go to the PLC's web interface and take screenshots of the Ethernet statistics page.

Just a guess but it is possible that the connections could be nearing the PLC's limit so it is dropping the FTV SE connections to handle others.

Refer to http://literature.rockwellautomation.com/idc/groups/literature/documents/at/enet-at003_-en-p.pdf

The PLC's and HMI(s) are all on the same subnet, yes?

brendan.buchan said:

By wire framing I'm guessing you mean the connection between PLC A and FTV SE drops out?

How loaded is the PLC, both the CPU and communication connections? Does it have a lot of MSG instructions? Can you go to the PLC's web interface and take screenshots of the Ethernet statistics page.

Just a guess but it is possible that the connections could be nearing the PLC's limit so it is dropping the FTV SE connections to handle others.

Refer to http://literature.rockwellautomation.com/idc/groups/literature/documents/at/enet-at003_-en-p.pdf

Click to expand...

Thanks for the replies.

rdrast the processor is on a different subnet. The crazy thing is it was working before then started gradually getting worse over the last few weeks.

When I say wire framing I do mean the connection between SE and the PLC drops out. But I can stay online and all the ethernet I/O works. I can ping the processor without any loss. I used the tool in rslogix5000 and only 24 out of 200 plus connections are being used. The PLC is on a different subnet. 255.255.255.224 my server is on a 255.255.255.0. The PLC that drops is on the /224 and i am within the available number of IP addresses which I believe is about 32. I would reboot the router but another processor is not having problems on this subnet to the server on the 255.255.255.0. Any changes require a plant shutdown pretty much.Flows are shut off etc.. So we don't get any violations.

I replaced the processor. I tried smart ports today and that didn't help. I built another server on the VMware and for the first hour the data was steady no drops on the new test server. Then the new test server pointed to [local host]. After about an hour started getting the same problem. I built this server with a fresh OVF template and installed FTV. Imported the project and entered all the enterprise shortcuts.

I tried smart ports but that did not work. I reset the OSI PI servers today as well.

The fact that I can ping the processor and stay online but the data is lost to factorytalk se goes against the way I always thought this stuff worked.

I am going to change all the fiber ports to 100 full across. I have a switch in one building that carries the network to the control panel with PLC A and PLC B. It has a port saying link lost and it flashes amber although some how PLC B is working with no problems. So in a nutshell I have a Stratix 8000 going to a 20 port 5700 for a PLC, then the network goes to another 5700 stratix 20 port switch in another building then goes to a router to jump onto my /224 subnet.

Another thing I might try is setting up a Rslinx classic data server to see if rslinx works just as a test. The other thing is install FTV patches. In addition we have a server at a different plant that is reading data from the same processor and I have yet to see that server lose data. I wonder if somehow the shortcut being named the same although on a different factorytalk directory could be causing confusion to the Rslinx enterprise at the plant having the problem. I might shutdown that server which is used mainly for status of all plants and see if the local plant continues to have data drop outs.