We are running 2 node Availaibility Group FC servers with 3 numa nodes. On our Prod servers the PLE's differ from node to node on the Primary node yet and are the same on the Secondary node yet on our Accp servers the PLE's are identical on each node of both the Primary and the Secondary Application-Intent Replica.
It might be expected to have different PLE's per node because of the workload. Could this also indicate some kind of targetting of a node by processes?
We are having performance issues on Production and this is also reflected by high write latency on TempDB which is on an SSD.
Any ideas?
Thanks
Chris