We were using nam3 and recently switched to nam6, we are observing 200 to 250ms increase in query latency. i understand we have to take some burden on going across different machines and different regions. But ~250ms increase is too much. Our app cluster in east-us, do we have any recommendations to improve the performance?
Moving from nam3 to nam6 changes your default leader region from us-east4 to us-central1 [1]. You may want to consider whether you can allow for greater read staleness or move your clients closer to us-central1.
"For optimal read performance outside of the default leader region, use staleness of at least 15 seconds." [2]
[1] https://cloud.google.com/spanner/docs/instances#available-configurations-multi-region
[2] https://cloud.google.com/spanner/docs/instances#multi-region-best-practices