Hello everyone! we ended up moving our backend (Spicedb client) to the GKE cluster where the Spicedb cluster is running, managed by the operator. It has been working really well (as you guys pointed, seems like the main problem was the LB service type, now we have as ClusterIP, the operator default) but we keep seeing the error

Error: 13 INTERNAL: Received RST_STREAM with code 2 (Internal server error)

now and then, but only in the client side, no errors in the Spicedb cluster pods.