Finally, I don't have any good way to check buffering on multiple nodes, other than checking the PI Server Network Management Statistics for connections from pibufss.exe.
Refer to the following KB article: http://techsupport.o...n/8/KB00363.htm
Q1. How do you monitor and support and upgrade 10 or 20 or more interface nodes?
Refer to the following KB article:http://techsupport.o...n/8/KB00363.htm
Well, you could sign an Enterprise Agreement and get Managed PI.
OK, maybe that's not an option. Let me start by asking what are your biggest challenges in managing your interface nodes? Is it an issue of upgrading? Monitoring data flow? Ensuring buffering is configured correctly?
If you need to monitor data flow and interface health there are a few tools that come to mind, like UniInt Health Tags (you might consider monitoring Message Count, Interface Status, and IO Rates). You can also use the Interface Status Utility (ISU), which monitors a single tag for data flow, if the data flow stops, it sets itself to an error status. From there it would be simple to create some PI Notifications to monitor your health and ISU tags.
When it comes to upgrades, I don't know of any way to automate the process. For every interface we are likely looking at a different data source that may respond differently to the upgrade. That said, having done hundreds (if not thousands) of interface upgrades, they work without any hitch at all in the vast majority of cases.
Finally, I don't have any good way to check buffering on multiple nodes, other than checking the PI Server Network Management Statistics for connections from pibufss.exe. If you want to check configuration you need to ensure that the /host parameter of every interface node exactly matches the buffered server configured for pibufss (it's in the piclient.ini file). In addition, you need to check that all of the interface services are dependent on buffering. I can see that it would be a tall order to try to accomplish that quickly across 10-20 machines.
I wish I had a better answer for you on this one. Have I at least identified the problem correctly?
Retrieving data ...