Firstly, right now we only can use PI SDK since we don't have AF installed in our PI System (hope soon we will have it). We are developing a tool that reads around 50k tags each 10 seconds or less, in order to get all new values from these tags. Our intention is read around 200k tags maximum.
Until now, we are achieving this using bulk reads "PointList.Data.get_Snapshot(out ptNamedValues)", splitting those 50k in PointLists of 10k each. I have noticed that sometimes we are missing one or two updates of a tag that shows up in the archive system when I check via Datalink.
My doubt is what is the best approach to continue with this project, keep using bulk reads or switch to EventPipes. Can EventPipes deal with the amount of tags?
Thanks for your help.