Using already posted suggestions the PLCs are running periodic tasks and the tag groups are at prime number intervals to prevent overlap (the leased one being 1123 ms). This is on 8.1.1 using the Ignition Logix communication driver for L8 PLCs running v32 and v33 firmware. We tried increasing the Max Concurrent Requests and CIP Connection Size in the device settings but the results surprisingly ended up being worse. The clients themselves seem to respond fine and don’t show signs of performance issues but we are afraid if we add more clients/PLCs that may not be the case anymore.
Any help on how to get the device load to a reasonable level would be appreciated and thanks for reading!
If the responsiveness isn’t a problem then you can probably just ignore the load factor being over 100%.
If you really want to chase after this then there’s a couple things you can try. The first would be to just reduce the number of tags you’re requesting. The second would be to request them slower. As an aside, there’s almost no meaningful difference between the 2 rates you’re using, so maybe just put them into a single tag group @ 1500ms.
More difficult things to look into are how your UDTs are structured, whether they contain any members that are not externally accessible, and whether or not you have tags coming from AOIs or not.
Edit: missed that one group is leased, that’s fine, you need a separate group for that. Staggering tag groups by prime numbers is just nonsense though. Ignition isn’t a PLC, these aren’t guaranteed scan times, it just doesn’t matter or make a difference. Requests belonging to each rate group will get all get queued to execute at once and then naturally settle into an execution “rhythm” based on all the other requests fighting to be executed.