Changing status.
Addition clarifications on use case - user wants to be able to designate a specific PI Collective member for retrieving the inputs for analyses, but write back the results using BufSS to fan the results to all collective members. This is currently not possible.
The PI analysis service is hardcoded to only connect to the primary collective member regardless of if you configure the AFSDK to prefer the primary. The reason being is that the analysis service needs to be able to create PI points. This means that if your primary collective member is not available, then your calculations will fail. We have a userbase that uses the output of analytic calculations for real-time monitoring. They are effectively blind when we perform any patching or if there is a network interruption between the analysis service and the primary collective member.
The analysis service should be able to read data from a secondary and create points on the primary as two separate processes. While the primary is not available, then the analysis service should queue those points for creation.
https://softwaresupportsp.aveva.com/en-US/knowledge/details/000092545?lang=en_US