I would like to understand a little more why PISDK is unable to fan data within a collective when writing values - is there any documentation on this specific subject?
I see there are limitations with a secondary PI system with SDK write values but that can be overriden...what I am getting at are applications that are constantly running and base their current computations based on data stored in a PI server during their last execution cycle. If there is a failover during cycles the application needs to know the last cycle values. The application can be "rewound" to a date/time in the past to recalculate forward in time if required.
PItoPI can be used but what would be nicer is for SDK to do the replication. Aware SSB is on its way but until that time is there any obvious issues if we wrote a data fanning component for data?
- Capture Updates to PI server. (Within application)
- Is PI server part of a collective? (Yes)
- ListMembers and check connection where Priority not -1.
- Check AvailableCollectiveBehaviors for each server to ensure PISDK writes available.
- Update each available PI server in collective.