very possible - we don't know for sure, and have never been able to induce the issue. The other instance of our spicedb client, running on another container in the same AWS EKS task, was able to communicate with the spicedb cluster just fine though, through the same LB, at the same time as the other one was having an issue. 🤷‍♂️ unsatisfying, but seems to reliably not be an issue now that we have a bit of grpc config in there, so great.