Real Time Snapshot Support
PI System Connector: Need real time snapshot data support for the PI System connector.

The issue identified with PI System Connector 2.3.0.1 has been resolved, and PI System Connector 2.3.1.1 is now available with snapshot support for PI Points that have compression disabled/off.
We are still on track for a follow up release in Q1 2021 that will expand upon the snapshot support introduced in 2.3.1.1 by adding configuration options and snapshot support for compressed PI Points.
28 comments
-
Joshua Duncan commented
In response to Matthew Bailey, "We need this as well. We have a project ..."
I wouldn't get your hopes up. I received an update on this on January 4, 2019 from a connectors product specialist at OSI. He reached out to the product manager for System Connector and they have decided not to support snapshot data collection with PI System Connector. On hearing this news my org has decided to remove System Connector and go with PItoPI/APS instead. The PI System Connector seems like a half-baked solution right now. I've heard from OSI that it should not be seen as a PItoPI replacement (see here) but instead should be seen as a solution for migrating AF DB structure from one AF environment to another. However, as of now, System Connector doesn't support migrating Analyses or Notifications. I have to say that System Connector seemed to handle PI tag creation/updates a lot more seamlessly than APS but also there was much less control and customization. With time I'm sure System Connector will become a solid product but seems lacking as of now. -
mbailey commented
In response to Matthew Bailey, "We need this as well. We have a project ..."
I should add that we have been waiting for the connector to meet these requirements for 4 years!
Essentially we want two collectives ( a source and destination) to be the same.
Requirements
- Send PI data even if it's not in AF.
- Send real-time snapshot data. It seems really useless for us to be able to only send archive data. Users using the destination system don't get the real-time or accurate values? A breaker is closed, but the users see's it open? Not making sense.
- If a tag name is changed on the source collective, it should update on the destination side.
- Should be much faster then PI to PI and APS (multi threaded), should work better across two locations where latency comes to play.
- Should be easier to configure then PItoPI and APS -
mbailey commented
We need this as well. We have a project on hold until this is done.
-
Joshua Duncan commented
I heard that this feature is in the works. Is there a tentative timeline for this? Like everyone else here, I agree that this is a very necessary feature.
-
Trent Huhn commented
Another vote for this functionality. We replicate data from remote site PI servers to a central enterprise server - having snapshot data available here to trigger real-time alarms is critical. Without this functionality, we are unable to implement PI System Connector at this time.
-
Laurie Dieffenbach commented
Another example: One of my customers has a situation where manually entered data or KPIs are calculated on a daily basis, or even hourly. The snapshot value won’t get to our central server till the next hour/day/etc. I don’t think there is a way to force the value to automatically go from snapshot to archive. PI always needs the current value and the previous value to decide if it needs to archive. Therefore, they need the snapshot value to be included in the PI System Connector.
-
Fabiano Batista commented
This idea is very important, Tood. Thanks for sharing.
I would like to mention is a typical use:
- Daily KPIs are calculated at the source system, and they should be available immediately on the destination server as soon as they get calculated in the source.
Unfortunately, with the current implementation of PI System connector, we see a lag of 1 day between the last daily result available in the destination PI System and the one available in the source PI System. One of my customers is having this issue right now. -
fburrows commented
A must have if replicating to an enterprise server that is then used as the primary source of data for end users.