First, sorry for my bad English, i hope nevertheless we understand us
Maybe you can Help me with this Problem, hopefully in the Portuguese Forum is it fixit.
So i have a MoellerEasy OPC server, he becomes Data from a W&T Comserver connectet to the PLC, ScadaBR connect to the OPC-Server.
All is still working, put the OPC Data only updadet once by restarting the Tomcat. With other OPC-Client (anyone) i can Read/Write the EasyOPC variables continuously.
Is it necesery/possible to update the opc-driver.jar to fix it, or did i have to change the Tomcat.
This issue was not yet noticed. For a fact, thanks to our French friends in Ville D´Albi, whom themselves developed another OPC Driver for ScadaBR, we shall have some improvements in next versions.
For now, I would ask to you to search in the log files for errors about OPC communication, and review your DCOM settings for OPC usage.
Most users coming across OPC challenge in ScadaBR, have issues with DCOM.
If you find anything suggesting errors in the logs, please report back, and we will try to help you through.
In mango.log, and stdout.log I found your problem (or ScadaBR´s).
You made the first poll in - 2012-07-02 20:45:50,343, couldn´t find some data points untill BeanConvert clashes with⌠itself??? duh
WARN 2012-07-02 20:45:50,328 (org.openscada.opc.lib.da.Group.removeItem:431) - Unable to find item âMoellerOPC.net02.circuit_diagram.S01â
WARN 2012-07-02 20:45:50,328 (org.openscada.opc.lib.da.Group.removeItem:431) - Unable to find item âMoellerOPC.net02.circuit_diagram.S06â
WARN 2012-07-02 20:45:50,328 (org.openscada.opc.lib.da.Group.removeItem:431) - Unable to find item âMoellerOPC.net02.circuit_diagram.S07â
âŚ
WARN 2012-07-02 20:45:51,500 (org.directwebremoting.util.CommonsLoggingOutput.warn:59) - Clash of converters for br.org.scadabr.OPCItem. Using org.directwebremoting.convert.BeanConverter in place of org.directwebremoting.convert.BeanConverter
âŚ
From here on, it trys to poll, and abort each time, saying it has a unfinished poll attempt⌠and at 2012-07-02 21:28:04,484 he crashes from lack of memory:
Exception in thread âpool-1-thread-35â WARN âŚ
java.lang.OutOfMemoryError: Java heap space âŚ
So⌠My guess: The OPC driver trys to make a poll, fails (due to ???), can´t close the thread properly and keeps building thread over thread till mem is full and tomcat crashes.
Quick cheat for this (here again, this is a long shot), raise the polling interval. This may give ScadaBR´s OPC driver enough time between polls to recover from any error.
Or wait for ScadaBR 1.0 with the advances in OPC driver. If we maintain the current OPC driver, be sure that your logs will be accounted for =]
where can i raise the polling interval, except the update period in OPC DA Properties.
This is now 2 seconds (befor it was 1 second) such as OPCServer and ComServer and tomcat also crash. The only different is about one more hour runtime of tomcat. But this, i think is the bigger memorysize ive gave the virtualmaschine. So i cant see any result changing the update periode.
Bewildering is that tomcat run for one hour, and then he begins to poll with the logical endresult OutOfMemory.
When did you await ScadaBR 1.0, or is it possible to become a beta?
You can mail me at squidy at the ftp serveradress from logfiles
I´ll ask for Victor to take a look at this. His knowledge in OPC goes way beyond mine.
We intend to deliver ScadaBR 1.0 by end of July, start of August.
