> The LE HTTP challenge gives no guarantee which A record it will use. That almost changed[0], but the current consensus[1] seems to be that you should be usin

> The LE HTTP challenge gives no guarantee which A record it will use. That almost changed[0], but the current consensus[1] seems to be that you should be using dns-01 for validation behind a load balancer instead. [0] https://github.com/ietf-wg-acme/acme/pull/138 https://github.com/ietf-wg-acme/acme/pull/138 [1] https://mailarchive.ietf.org/arch/msg/acme/6RSxtvpkcQNPmKs9f45k_bi9QVo https://mailarchive.ietf.org/arch/msg/acme/6RSxtvpkcQNPmKs9f...