We configured a load balancer this afternoon during the low-traffic window in preparation for implementing a clustered configuration of the community application. Behind the load balancer is still a single node, so visitors won't see any changes in performance or stability, but the next step is to implement the second node in the cluster, which will bring improvements. Meanwhile, there is a known issue that causes an occasional error when viewing threads. The error resolves itself when the page is refreshed, and it will be fixed when the cluster is complete.