I like the fact that we can now schedule archive shifts. We plan to set ours to shift daily. My understanding is when the individual collective PI server members time hits the scheduled archive shift time, the new archive is created. We feel this is useful so we can copy archives within a collective to other members if need be and not need to manipulate the start and end times of the archive files.
The issue I have is if the time is slightly off between collective nodes, there will be a slight data gap. I was wondering if anyone else has seen this or shares this concern?
I think the best way to handle archives shifts would be to have the primary node initiate them at the same time across all collective members, despite the local time of the other servers.