Looks fine I think. [1][2] # soa for s in $(seq 4);do dig +noall +answer @ns"${s}".google.com gmail.com;done gmail.com. 300 IN A 172.217.14.197 g

Looks fine I think. [1][2] # soa for s in $(seq 4);do dig +noall +answer @ns"${s}".google.com gmail.com;done gmail.com. 300 IN A 172.217.14.197 gmail.com. 300 IN A 172.217.14.197 gmail.com. 300 IN A 172.217.14.197 gmail.com. 300 IN A 172.217.14.197 # cf open resolver dig +noall +answer @1.1.1.1 gmail.com;dig +noall +answer @1.0.0.1 gmail.com gmail.com. 293 IN A 172.217.4.69 gmail.com. 144 IN A 142.250.69.197 # google open resolver dig +noall +answer @8.8.8.8 gmail.com;dig +noall +answer @8.8.4.4 gmail.com gmail.com. 300 IN A 142.250.69.197 gmail.com. 300 IN A 142.250.69.197 Perhaps it was a transient issue at CF? A historical monitor like thousandeyes might have seen it. It gets tricky to catch with Anycast memberships. IIRC each of their Anycast regional groups at CF have a dozen or so nodes but there are several CF employees on HN that could correct me. [1] - https://www.whatsmydns.net/#NS/gmail.com https://www.whatsmydns.net/#NS/gmail.com [2] - https://www.whatsmydns.net/#A/gmail.com https://www.whatsmydns.net/#A/gmail.com