This has been a minor annoyance for me (if it was a big problem for you, why didn't you say something earlier?) but with the Dnstap Reloaded work going on in ShoDoHFlo on the fwm branch it's gotten a lot noisier.
The fix is on the fwm branch.
The technical issue concerns empty non-terminals, FQDNs which resolve but have no answer of the appropriate type. Due to the work to support SVCB and increasing popularity of 6+4 "happy eyeballs" the errors are becoming much more frequent.
They've never actually represented a problem for Rear View RPZ: if there's no address then the event is ignored. They were logged because it was thought they might reflect an issue with the upstream telemetry.
If you're running ShoDoHFlo off of the fwm branch, then run this off of the fwm branch too. Eventually this will all get merged into main.
This has been a minor annoyance for me (if it was a big problem for you, why didn't you say something earlier?) but with the Dnstap Reloaded work going on in ShoDoHFlo on the
fwmbranch it's gotten a lot noisier.The fix is on the
fwmbranch.The technical issue concerns empty non-terminals, FQDNs which resolve but have no answer of the appropriate type. Due to the work to support
SVCBand increasing popularity of 6+4 "happy eyeballs" the errors are becoming much more frequent.They've never actually represented a problem for Rear View RPZ: if there's no address then the event is ignored. They were logged because it was thought they might reflect an issue with the upstream telemetry.
If you're running ShoDoHFlo off of the
fwmbranch, then run this off of thefwmbranch too. Eventually this will all get merged intomain.