PLC Connect Aspect RealPCA.value goes Bad
When PLCConnect aspect RealPCA.value goes bad, the only remedy is restart of Aspect/Connectivity Server.
Kepware OPC Server resides on separate computer node "outside" 800xA. When that node is restarted problem occurs.
Why doesn't ABB OPC DA Service Provider "see" when it's back "up"?
Values are "Good" in Test OPC Client (ABB internal) or other OPC Client.
System version 126.96.36.199, Windows 2012 R2 Servers.
Voted best answer
The ABB does see that the Kepware Server is back up, **BUT** it is not getting a new value for the data - ie the data refresh doesn't happen. The data Should in theory start updating again whenever a new value appears in the Kepware server, which triggers a refresh. Of course, with a lot of your data this might not happen for a while. So the only way to refresh the values in the 800xA RTDB system immediately is to restart the PLC Softpoint Service again. This kills all of the OPC subscriptions to kepware properly and makes a new subscription.
That's why your other test clients seem to work - they made new subscription groups.
Sorry, not a great answer, but thats just how it is. Our experience is that you really do require Redundant PLC Softpoint services so that you can restart them as required - especially when you are making configuration changes. The Kepware OPC Server is normally very reliable and should not fail often. However, using a second redundant Kepware OPC server may help.
Typically the biggest problem is running tunnelling between OPC servers in different domains or workgroups, which is why my personal preference is to avoid tunnellers at all costs.
An OPC server that over time is ”forgetting” subscriptions that it once accepted may drive this behaviour.
This occasionally happen also when using OPC DA Connector to connect directly. Then, the server’s incorrect behaviour can be proven using applog (subscription is added but never cancelled).