AnsweredAssumed Answered

OPCInt Health Points error: Error reading system digital state

Question asked by cezary.kosla on Apr 5, 2017
Latest reply on Apr 5, 2017 by cezary.kosla



I am seeing a strange error in one of the OPCInterfaces I am working with:

The PI Server is reporting that the Health Points associated with the Interface have a value of: Error reading system digital state

While the Interface itself (if I check through ICU) is updating the Health Points with new snapshots.


How this happened was:

I had a working OPCInt interface (setup with Health Points and Disconnected Startup) which I had to ugrade into a Failover configuration with a backup OPCInt.


I configured the failover on both of the interfaces, but during the setup on the primary one (using ICU) I got a pop-up information (after pressing save) that ICU has to update the Health tags of the interface. I pressed No at that time as I still had some other options to put in into the Failover setup, when I finished with the remaining configuration options for the failover I pressed Save again but this time the ICU did not present me a pop-up mentioning the Health Points.

I restarted the interfaces both primary and secondary and checked that failover is working and new data is passed to the PI Server.


Next day when I was checking for stale points on the server I noticed that the Health Points from the Primary interface are reported either as BAD or STALE.


So I went back to the primary node, deleted the Health points and recreated them assuming this will fix the issue. As the interface is running under Managed PI service I also restarted the PDS service, accordingly to the instruction which I found in KB01231


But now.. all of the Health tags for this interface show something like this:

Health Points_digital state error.png


After deleting/creating the health tags I did not restart the interface, can this error be caused by the fact the interface is setup into disconnected startup?


I noticed that the Digital States cache file last modify date is yesterday at the time when I restarted the interface, so should an interface restart (and cache update) fix the issue?