Description of problem: The default configuration shipped with unbound has ipv6 enabled. Unless you are one of the lucky few that has ipv6 configured your syslog will get spammed with notices related to unreachable ipv6 addresses. Since people using ipv6 presumably know what they are doing, disabling it in the default config seems perfectly reasonable in the next few years. Version-Release number of selected component (if applicable): unbound-1.2.0-4 How reproducible: always Steps to Reproduce: 1. install unbound 2. start it up 3. resolve something connecting to localhost ($ dig @localhost www.redhat.com) Actual results: lots and lots of lines in /var/log/syslog for each unbound request of the following pattern: May 20 13:42:57 viktor unbound: [16389:1] notice: sendto failed: Network is unreachable May 20 13:42:57 viktor unbound: [16389:1] notice: remote address is 2001:503:231d::2:30 port 53 May 20 13:42:57 viktor unbound: [16389:1] notice: error sending query to auth server; skip this address May 20 13:42:57 viktor unbound: [16389:1] notice: error for address: 2001:503:231d::2:30 port 53 Expected results: not being able to connect to ipv6 hosts should be the norm, and the syslog should be quiet about that Additional info: I have found that adding "do-ip6: no" to the appropriate place in /etc/unbound/unbound.conf resolves this
This is being addressed upstream. It's either fixed in 1.2.1 or 1.3.0.