>>Message imported from previous forum - Category:ClearSCADA Software<<
User: dialvec, originally posted: 2019-06-25 20:57:28 Id:459
I am deploying a M2M network to communicate some remote RTUs. I currently have 3 RTUs connected to a central server.
I am using GPRS modems and I am time-randomly getting the following alarm message:
"Scanner Command (Primary Poll) Timeout"
The network has been configured to communicate every 5 minutes and modems transmit to each other via serial .
Also, network's connection is intermittent, so I get communications state Fail messages.
What kind of parameters should I check in order to detect the system's fault?
Thank you so much
Solved! Go to Solution.
>>Responses imported from previous forum
Reply From User: JesseChamberlain, posted: 2019-06-26 03:40:11
Sounds like you're doing the primary poll for all three at the same time, likely the time it takes to complete two of them is more than the timeout (30s) sometimes. Try offsetting the polls.
You should also review your polling regime, doing a primary poll every 5mins is very inefficient and doesn't take advantage of DNP3's features - your current setup is probably less efficient than modbus!
Reply From User: dialvec, posted: 2019-06-26 12:36:51
Thank you so much for your answer.
Could you please explain more about what settings could I use in order to take advantage of DNP3?
Once again, thank you
Reply From User: JesseChamberlain, posted: 2019-06-27 05:36:42
It's a bit of a big topic for detail on the forum, we run multiple day training sessions on DNP3! It will also depend on the feature set of the RTUs.
You will want to analyse the data you're gathering, how much it changes and how often you need to know. On the RTU, set event deviations based on how much of a change you want to know about. Classify your points into classes so you can poll or get unsoliciteds at different rates.
You might have Class 1 points set to send an unsolicited as soon as they deviate by 0.1 and a class poll of 5mins, but Class 2 points set to create events when they deviate by 10 and to only gather the event data hourly. It's a combination of the config on the device and the CS outstation config.
Reply From User: dialvec, posted: 2019-06-27 12:40:10
Ok. I get the idea.
Thanks again, you have been very helpful