[mtodd]

We have an identical load balancer to one that is causing problems so we're lucky enough to reroute traffic through that one instead to get to the same boxes. (The boxes serve two different APIs through two different DNS CNAMEs so we split the ELBs for future and sanity). In this case, it's helped us out. Alternatively, we would've just routed all traffic to our west coast ELBs.