View Full Version : Process Gateway working status in Runtime

18th May 2020, 09:02
Hello,<br><br>First of all, to give my question some context, this short introduction:&nbsp;<br>Very often, in energy industry specific applications we do, there is some kind of telemechanic communication channel with parallel or parent system.&nbsp;<br>This channel is of course created using Process Gateway.&nbsp;&nbsp;It is vital that it works continuously, but there were few cases of PG.exe failure. Since there is no (known to me) way of informing Runtime user about PG.exe issues, here is my question:<br><br><strong>Is there any way to show the status of every PG.exe in Runtime? </strong>Currently, I have VSTA macro checking every 30 seconds if process name containing "procgate" is currently working in Windows, but this solution is far from perfect if we use more than one PG instance - those cannot be differentiated by process name, which is the same for every instance.&nbsp;<br><br>I hope you can help me solve this problem.<br><br>Best regards,<br>Adam<br><br><br>

18th May 2020, 10:25
The way to make&nbsp;a "watchdog mechanism"&nbsp;checking if the (independent) Process Gateway application is still running and responding correctly -&nbsp;depends on communication protocol.<br>&nbsp;<br>For example, in zenPG IEC870Slave via TCP&nbsp;(IEC 60870-5-104)&nbsp;I would recommend to create in zenPG configuration an additional 'device' for 870-master with IP of the zenon Network process server(s) and&nbsp;in zenon project to add the iec870 (master) driver - connecting this 'device' in zenPG. <br>In zenPG there can be 2&nbsp;IOs, one&nbsp;T30 and one T45&nbsp;linked with zenon&nbsp;INTERN variables (BOOL).&nbsp;In the project the&nbsp;'Allocation' can assign the INTERN variable received via T45 - to INTERN variable sending T30. With 'Time Control' module you can&nbsp;trigger periodically (e.g. each 30s) 'set value' functions linked to the iec870 (<u>not</u> <u></u>intern)&nbsp;variable&nbsp;T45 - to send the command to the zenPG.&nbsp;The T30 variable of iec870 - with then received&nbsp;the value change - you can link with reaction matrix with <u>delayed</u> <u></u>alarm, with delay&nbsp;40s (30+10).&nbsp;Then, the&nbsp;alarm will&nbsp;appear when&nbsp;there&nbsp;is no periodic value change - means either zenPG is not reacting on commands or not able to send T30 update.<br><br>In this zenPG there is&nbsp;also T00 with IOA 8 to detect the start of the process (or reload /restart&nbsp;it) - common for all&nbsp;devices.&nbsp;And, pro device,&nbsp;there is T00 with IOA 1 "master connected" information.<br><br>Anyway, I would strongly recommend to sniff and examine what the control center (parent system) sends to zenPG making it hang of crash. Probably the remote system is producing some fuzzy frames.