Ignition OPC UA Timeout and Failover Settings

Ignition
1

Hello, I was experiencing a faulted OPC server connection on my redundant backup Ignition gateway. The status was fine on my primary server. I had found another thread where the user was experiencing a problem with a faulted connection because the OPC endpoint had failover enabled, but there was no failover endpoint URL specified. I checked my settings and noticed that this was the case on this system as well. I disabled failover on the endpoint since there was no failover endpoint URL and the connection status was now successful on both the primary and the backup gateway. I logged off the system and about 15 minutes later, I received an email from the client with a picture showing a connection loss.

IMG_1630
IMG_1630480Ă—640 78.4 KB

I logged back into the system and by this time, all the connections were showing successful again and the client informed me that everything was back up. I went to the error logs and saw that there were timeouts on the OPC connection that I had disabled the failover on just a few minutes after I had done so.

MicrosoftTeams-image (2)
MicrosoftTeams-image (2)1294Ă—791 59.7 KB

Apparently, this isn’t the first time the client has lost connection, but it doesn’t happen often, and it seems quite suspect that it happened just a few minutes that I disabled failover on the exact OPC connection that would go on to timeout. Does enabling failover without specifying an endpoint change the behavior of the connection? Assuming there is an issue with the OPC endpoint, would this failover potentially mask the issue by causing more connection retries? Also, the image I received from the client shows “Gateway Connection Lost” and not a message specific to a single OPC connection. Would an OPC timeout cause this message or was there a separate issue that caused a connection loss to the entire gateway?

Thanks!

2

The screenshot of the client has two critical pieces of information:

  • Both master and backup gateways are unreachable from the client–OPC is utterly irrelevant.
  • Both gateways report “no route to host”. That is a network failure.

This tells me your gateways are on an unreliable network, and lost connection to both clients and devices in the same event.

1 Like
3

Thanks. So changing the settings in the OPC endpoint could not have caused any connection issues between the client and the gateways? Also, I am curious if you have any idea about whether or not enabling failover without specifying an endpoint could change the connection behavior of that endpoint.

4

No. The clients talk to the gateways on http or https to the main ports. OPC is on separate ports.

No idea.