Is there a way to replicate batch across collective?
The batch interfaces will not create two sets of batches on a collective (no fanning on batch data). Also, the server assigns the GUID of the batch and they will be different. You can change a tuning parameter on the PI Server and that will allow you to write batch data to the secondary; however, the batches will not be copies of one another. Programs like RtReports will not work with this and you will end up with duplicate reports and essentially a very rough road ahead. There are programmatic ways to assign the GUID and I can recommend people who made tools around that; however, you probably would not want to run it in real time for your electronic batch records. It is interesting, that over the years I have heard very little on HA from life science customers. Nearly all were more concerned about data integrity and ability to mount backups than on having high availability. I know quite a few have HA and are running batch on the primary and some have switched or will be switching over to Event Frames. The batch information can usually be recovered from the BES/MES and any PI data during downtime can be be buffered.
Overall, I would never recommend a life science customer trying to use the batch database on the secondary server. The best approach is to use Event Frames with a technology such as SQL Always On. You will have to check with OSIsoft, but I understand RtReports with Event Frame support will be awesome and coming out soon!!
Thanks for the great response, Dan. Roland, I would echo Dan's comments about not implementing any batch replication onto the secondary PI Server, and that HA is recommended if and when you switch from batches to event frames, at which point you have a number of options including SQL Always On.
Retrieving data ...