From 8:00 PM until 8:35 PM there was a single node in our Dallas cluster which issued "550 5.7.1 Command rejected" to incoming emails from the entry loadbalancer. The rejection came from a health checking system on the node which took the out of service as it detected a fault with its ability to resolve DNS entries. The rejection was correctly issued, however, it was issued with the incorrect status code which ultimately lead to a rejection of the incoming message instead of a temporary rejection. We have corrected the behavior of the health checker across all nodes to issue a temporary rejection code instead of a permanent failure.