Hi. We have a cluster with 50+ nodes. When we redirected app connections to the ZDM Proxy, it started reading from only 8 nodes, and all others dropped to 0 reads.
I didn't find anything helpful to understand this behavior. Is there anything I have to change?
We also had an issue with writes, but after a period, it was sending writes to all nodes (apparently)
At 14:30, we rolled back and moved app connections directly to cassandra.