|
|
Rank: Advanced Member Groups: Member
Joined: 7/16/2008 Posts: 35 Location: UK
|
Hello, Having a nightmare here just now with the OPC DA Interface....OSI are scratching their heads over this one too...the OPC server appears to be healthy, we can use the OPCTool from OSI to see tag data and we see good data (00C0)....but the PI interface is just reporting bad for every tag. We put a debug on a tag and it thinks the data quality is bad. Tried a few more 3rd party OPC clients on our API node and they could all read this OPC server (and all figured the data to be good)....succesfully worked our way from the old 2.1.x interface to 2.3.x...  Had the control engineers reboot the OPC server a few times, nothing has changed in terms of DCOM etc...originally a few tags were reporting "bad" we restarted the interface and now every tag is "Bad"....anyone seen anything like this? Who ate all the PIs?
|
|
|
|
|
OSIsoft vCampus is a subscription-based, online offering that consists of providing everything people need to develop applications on the PI System. We invite you to take a "tour" of the OSIsoft Virtual Campus - also feel free to consult the FAQ or contact OSIsoft vCampus for more details.
|
|
|
Rank: Administration
 Groups: Administration
Joined: 6/20/2008 Posts: 612 Location: Cheshire, United Kingdom.
|
It does sound like security or even asynchronous callback is failing. Are you sure nothing has changed? Usually some control engineer changes something without telling anyone...  If OSI are stumped then it sounds like a tricky one. Anything in the usual places, e.g. pipc.log files on each node??? When you say Bad, do you get any additional info about the Bad status? Principal Consultant Real-Time Data Management @ Wipro Technologies
|
|
|
Rank: Advanced Member Groups: Member
Joined: 7/16/2008 Posts: 35 Location: UK
|
Things have gone from bad to worse, now we can't connect to OPC server, I unloaded all tags and stopped all interfaces and left it to timeout any connections in the OPC server in the hope that the OPC server would give up and we could reconnect, but now when i try to connect to the OPC server from the OSI OPCTool the tool just hangs. It connects fine to OPCEnum to browse the server list so I don't think theres any problems in the DCOM, as this has happened over the xmas period, there wouldn't have been anyone there to interfere with the server (as much as we all know control engineers love making tweaks here and there!!!!) in fact no one on site (offshore) really has any experience with the OPC server (which isn't helping!) so I'm fairly convinced no changes have been made. It baffles me how we connected and all looked ok from Matrikon Explorer, OPCTool but the interface didnt like the data, now nothing will connect......  re: the bad status, when we debugged some tags, in PIPC log we saw "Bad Quality, non-specific" ... in the OPC explorer, "Good, non-specific".... I'll keep you posted if anything amazing and new comes out of it (other then the more likely, "it just started working again AKA whatever changes they made were reversed") ;) Who ate all the PIs?
|
|
|
Rank: Member Groups: Member
Joined: 1/14/2010 Posts: 28 Location: UK
|
i dont suppose the firewall rules have changed?
|
|
|
Rank: Administration
 Groups: Administration
Joined: 6/20/2008 Posts: 612 Location: Cheshire, United Kingdom.
|
Pho wrote:i dont suppose the firewall rules have changed? Good point (welcome to the forum!) It is usually the simplest thing and sometimes the obvious. When something is working, then stops working something has to have changed. Wonder how he is getting on, I bet he has had a few sleepless nights after trying to solve that problem. Principal Consultant Real-Time Data Management @ Wipro Technologies
|
|
|
Rank: Member Groups: Member
Joined: 1/14/2010 Posts: 28 Location: UK
|
Thanks! I can see me being on here quite a bit :-)
We had a similar issue a while ago and luckily that was the first thing i checked as we too have "others" who play with things which effect us. Its all good fun :-)
|
|
|
Rank: Advanced Member Groups: Member
Joined: 7/23/2008 Posts: 38 Location: Cheshire, UK
|
Hi MacoroniPi,
Did you manage to resolve this issue. OPC is a forte of mine, having been certified to Level 4 (highest level) of the certified OPC Professional qualification.
The most obvious place to start is with the pipc logs on the interface pc. Then try connecting with opc browsers from a remote pc i.e. the PI server.
One of the things I have found in the past is that the IT departments seperate network domains, etc, and this impacts the security for launching the interface and is typically seen in the pipc log on the interface box. The fact that things worked before and now don't work points more to an external influence on the system rather than a PI issue.
What other Interfaces do you have other than OPCDA and do these still work? What are the domain specifics in use? How is your interface setup, to read everything as a string, or as values.
Any additional information you can supply would be useful in determining the root cause of the problem.
Regards
|
|
|
Rank: Advanced Member Groups: Member
Joined: 7/16/2008 Posts: 35 Location: UK
|
UserInt1 was the culprit, this had a value of 1 on all tags, which was attempting to load the value as an array. For some reason we did not see this in the PIPC log until I had installed an OPC repeater to relay the values from the OPC server to the PI server from the PI server itself; as we were seeing good data in the OPC repeater but digital state BAD in PI we were able to rule out any DCOM issues etc and then spotted the data mismatch errors in the pipc log. No one has (knowingly) changed the userint1 field, but when we changed it to 0 and switched back to the main OPC server, it started to receive good data and we were back up and running. We use templates for Excel for tag manipulation work (PI-SMT) and do not have UserInt1 field in this template so all the fields we checked between the tags which weren't working and the one or two tags which did work all looked ok. In the middle of this problem being started and being resolved the OPC server went down for 2 weeks (with no one in the production side being able to do anything about it as only one person knew anything about the software and he was on holiday!) so that didn't help either  Who ate all the PIs?
|
|
|
Rank: Administration
 Groups: Administration
Joined: 6/20/2008 Posts: 612 Location: Cheshire, United Kingdom.
|
Glad you got to the bottom of it - certainly worthwhile information for anyone with similar issues! Principal Consultant Real-Time Data Management @ Wipro Technologies
|
|
|
|
Guest
|