"Point value may not be reliable"
-
Hi guys!
I have this error on the view on a DataSource and on all of the DataPoints in this DataSource. The yellow exclamation marks are on view next to the values.
My problem is: I don't know what is the issue. The Mango has stable and fast connection to the DataSource, I tested multiple times. No messages in the log (stdout), no alert in the Event list.
I can see the precise values on the view, no aborted poll, nothing. Everything is working as expected! If I press the refresh value next to the warning, it runs and the new value will show up as I expect.
Can you help me guys how to hunt down this issue? The users are bugging me to remove this yellow exclamation mark from the view.
Thank you!
-
Hi gbodacs, welcome to the forum!
Hmm. I would have expected an event, but I guess it varies based on the data source. What type of data source are you using?
-
Hi Phil,
I use ModBus protocol on an Adam 6015. The other Adam is on the same router has no problem at all (well, no "point value may not reliable" warning).
I tried to "restart" the connection between the Mango and the DataSource, changing a setting and save it. I can see in the Mango log (stdout) that the communication restarted and it took ~1300 msec. After this "communication restart", the warnings are going away, no warnings on the views. After that, the first automatic refresh (it set to 15 minutes), the warnings are coming back. I checked the log and the events for this time period and found nothing. What do you think, do I have to go back in time further to find the event you expect?
-gbodacs
-
Looking at the code, modbus data points are marked unreliable after being enabled before a poll. They're also set unreliable if there is an ExceptionResult (like reading an illegal address, meaning the device responded something was in error or didn't respond with all requested data) or if there is a timeout exception or other transport exception. Any successful read for a point should clear that unreliable key.
In my experience this is usually caused by an illegal address response, which should have created an event if your data source has its default event levels.
You can generally resolve it with either a decrease in the max read registers / bits, or by checking 'Contiguous batches only' which will only read the registers there are points configured for (as opposed to trying to minimize the number of requests.
-
Thank you for the detailed answer!
I checked everything you mentioned in your answer and I have found the problem!
"or if there is a timeout exception" <-- that was exactly the problem! The timeout had been set to a wrong value (a 0 was missing at the end of the number).
I successfully reproduced the issue with setting the exception events to "Urgent". I got the warning and an alert email too with the timeout exception, so I'm pretty sure I have fixed the issue.Thank you for the help!
-
Certainly! Glad you got it resolved. Be sure to keep an eye on the data source event levels in the future while troubleshooting!