I could not really find any good references on this topic, so let's ask around.
If we were to place the PI Servers (DA, AF) remotely, to what extend will the network latency cause problems?
What i think i know:
- For interfaces. this does not really play a role (besides a KB on UFL)
- For synchronisation between geographically spread PI Servers this seems manageable (there's a KB on that)
- Some applications, like PSE and especially DataLink, from reports i've heard (so nothing really nicely measured), tend to suffer in usability when the latency goes over 20ms-30ms
Any of your experiences? Would a latency of 200ms be an issue?