Ignition Edge OPC Server Disconnections

Hey all,

I’m working on a project at the moment where I have a Siemens S7-1500 PLC hosting an OPC server that an ignition edge device reads tags from. This setup has been working for a few weeks now, but after a hard restart today we are are having an issue where the OPC server is disconnecting and reconnecting at about 2 minute intervals.

I manged to grab a screenshot of the logs when the OPC server faulted, but didn’t manage to fix the issue based of this.

Screenshots of the process I followed to setup the OPC client is below.




So far I have tried restarting the PLC and edge device numerous times, recreating the OPC server and client, rolling back the gateway to a previous working version, connecting the PLC directly to the edge device via an ethernet cable, setting up a new trial version of ignition as an OPC client (this managed to connect and fucntion correctly), tried on both ubuntu and windows versions of ignition edge, confirmed that the edge device could ping the PLC, temporarily disabled firwalls and securtiy.

If anyone had any suggestions of other things to try that would be greatly appreciated. Thanks in advance.

It doesn’t look like you’ve done anything incorrectly in Ignition. This server is just periodically not responding to keep alive requests or closing the connection for some reason.

How many tags are you subscribed to on it? Are there other OPC clients connected to it? Can you check the CPU/comms load of the PLC?

Hey Kevin,

Subscribed to around 7500 tags, maximum on the PLC is set to 10000. The ignition edge device is the only client on the OPC server. Not sure about the comms load, but the PLC was running at about 25-50% when ignition was connected, then spiked to about 75% as it disconnected.

If you can use Wireshark or tshark to capture traffic between Ignition and the PLC it might be interesting to see if there’s any particular request or other pattern that is occurring before it closes the connection.

Came back in this morning to run these tests and it looks like the problem has managed to sort itself out. I’m thinking network traffic may have been the issue.

Thanks for the suggestion, I’ll keep it in mind if the problem comes up again.