On 6/9/21 10:50 AM, George Thessalonikefs via nsd-users wrote:
I think you posted on the wrong mailing list :)
oops! yep.
I'll reply here and include unbound-users also.
If you want to reply further please drop nsd-users and let us continue the
discussion only on the unbound-users mailing list as to not generate noise here.
thx.
On 07/06/2021 21:31, PGNet Dev via nsd-users wrote:
On my server, I config
what's the order policy for those forward-addr's?
Is the forward-addr used randomly, or sequentially, selected from that list?
Randomly between the fastest servers (in the <400ms band). Unbound will keep
probing the other servers to keep updated metrics for them in case they become
fast enough (<400ms). You can read more about it at
https://www.nlnetlabs.nl/documentation/unbound/info-timeout/.
Is there any unbound preference for IPv4 vs IPv6 forward-addrs?
There are the 'prefer-ip4:' and 'prefer-ip6:' options if you want to do that.
in the case of
do-ip4: yes
do-ip6: yes
prefer-ip4: no
prefer-ip6: yes
iiuc, then, it's strictly a preference, and _will_ fallback to the ip4 if ip6
'pref' fails ?
Reading for 'prefer-ip4',
" Useful if the IPv6 netblock the server has, the entire /64 of that is not
owned by one operator and the reputation of the netblock /64 is an issue,
using IPv4 then uses the IPv4 filters that the upstream servers have."
tbh I've no sense for the magnitude of that issue.
I'd guess it's a non-zero-sized problem. But generally _worth_ setting a
preference?
Are the settings' defaults (prefer-ipX = no) chosen as typically *recommended*
-- vs simplest config -- values?
Does the preference respect /etc/gai.conf?
No.
thx o/
