Hello,
one of our opc-ua servers restarted and Ignition was not able to reconnect and the connection was faulted. We were able to reconnect manually by disabling and enabling the connection.
We don't know exactly the reason, but it seems that the server had some 'ramp-up time' and during that time a connection to the OPC-UA server was not possible. Afterwards it works.
Is there a way to either increase the reconnect tries or monitor this via a script and start the reconnect again if the connection stays in faulted?
There are also no good logs for the reason. Would be good to have some more detailed warning/error logs if a connectes is eventually faulted.
Try to get a thread dump if this happens again.
It might have been this issue: OPC UA connections become unstable in Ignition 8.3.8 - #3 by Kevin.Herron
Connections don't (or aren't supposed to) stop retrying.
Will try to get a thread dump, although I hope it will not happen again 
I saw the other impressive post. There are some differences, but maybe the same root cause:
- We are on 8.3.6 | They are on 8.3.8
- We have only 4 outgoing OPC-UA connections and 1 incoming | The other post has ~200 outgoing OPC-UA connections.
- The OPC-UA Connection itself is stable as long as the server is available | The other post has an issue that "After some time, individual connections started entering the following error state...". Although maybe this is because of a disconnects due to an unstable WAN connection.
- The OPC-UA connection was successful, after we disabled the connection and enabled it again. | A full gateway restart was not required as it is stated in the other post.
- No OPC-UA thread is blocked (now), but also resolved the issue by disabling the and enabling the connection.
- We also have custom OPC UA data structures exposed by the server
These loggers may help:
- com.inductiveautomation.ignition.gateway.opcua.client.ClientManager
- com.inductiveautomation.ignition.gateway.opcua.client.connection.OpcUaConnection
- org.eclipse.milo.opcua.sdk.client.SessionFsm
- org.eclipse.milo.opcua.stack.client.ChannelFsm
and Wireshark is always the best as long as encryption can be disabled or reduced to Sign only instead of Sign and Encrypt.
You can monitor or set up alarms in the system tags for the OPC. I have seen it freeze due to a server restart also but had to restart the server again to get it going after doing the procedure you mentioned. I did verify the firewalls and the port it was connecting through were working properly. I also tried stopping and starting the Ignition service on the server first. I had a tag provider freeze up on me due to an alarm on a tag looking for users in a gateway notification list that wasn't configured correctly before also.
