I guess, for now, it is dns-forward-max hit. And the counter never gets reset and/or the connections to upstream never get timeout/flushed because the REFUSED reply lasted much longer than 40s(the timeout in source code) while no request going in for that period.
On Sat, Oct 19, 2024 at 9:54 PM チュimoc <[email protected]> wrote: > `dig foo.bar @127.0.0.1 -p 53`, 0ms instant return which upstream is at > least 10ms away. > But dig upstream the result is correct. Restarted dnsmasq it'll work for a > while then error again. I'm on the latest git. Running a debug screen for > now and want to get more info from dev. >
_______________________________________________ Dnsmasq-discuss mailing list [email protected] https://lists.thekelleys.org.uk/cgi-bin/mailman/listinfo/dnsmasq-discuss
