Is there any chance probability that data being monitored and logged on two Oprator stations be different.
My Freelance Network has a set of redundant controllers, two operator stations and a Server Machine. What i understand is that the operator stations take the data directly from the controller for monitoring in graphic dispaly and for data logging on the operator stations.Both are 100% redundant.i.e all displays are avaiable on both operator stations
I have configured data same data logging on both the operator stations for redundancy purpose. Which is tranferred to Server machine periodically throgh ftp protocol. My concern is about data consistancy..i.e can the data being monitored on both OS be different by some effects of EMI or so ...
and can data being logged on two OS and archived on server be ever inconsistant..
If no why? AND if Yes why and how it can be avoided.
can the data being monitored on both OS be different
by some effects of EMI or so ..
- No, ( well, maybe in some vague theory, but No. Lets be sensible here)
Anyway, yes, it is possible to get different data on two workstations when you use redundant OPC servers - usually because one connectivity server is faulty and one connectivity server is working correctly. This effect is rare, usually obvious to an engineer from the system diagnostic messages and easily fixed by rebooting the affected connectivity server. I have however seen this effect confuse operators who have not recognised the error messages.
You could also get the much more common effect that while both servers are working correctly they are logging data at very slightly different times. This means your data recorded on each server is "close", but not identical. However, in my exprience, the standard 800xA history logging does not seem to have this problem ( ie both servers should always log the same data).
In your case I would be more concerned about using ftp. This isn't part of the standard 800xA system and doesn't have the diagnostics associated with it that makes it compeltely obvious that some part of your software has failed.
Making sure that your data is logging correctly and is consistently is part of your daily maintenance and monitoring. The 800xA history system is quite reliable when properly configured and maintained.
In freelance as far as you are using Structure node or graphic display pool for monitoring on all the operator stations-DigiVis monitoring values are identical. There is possibility of inconsistency incase at anyof the digivis station display cycle time changed i.e default is 1sec if someone want fater cycle time user can go upto 200 ms, and in that case values are updating as fast as cycle time there can be inconsistency(of fraction of ms) between two different Operator Station.
Incase of archiving again if you are accessing common logs and trends in time synchronised system than data logged identically. only possibility is if there is no time synchronisation and values are not archived identical for particular time instance. else I don't see any reason for any inconsistences.
As per my understanding data in freelance transport via DMS protocol over Ethernet from controller to Digivis and again cyclically bundle of archive data is transferred to server using FTP over Ethernet in your case. i.e source of data is same and if systems are time synchronised there can not be any issue of inconstancy.
Regarding effect of EMI or so i don't think so as mostly measures are taken to avoid any interference between network wiring and power cabling. as well data transfered here on Ethernet cables (cat twisted cables).
Hope this gives some light on what you are trying to explore on this topic.