On 11/03/16 04:15, Stuart Henderson wrote:
On 2016-03-08, Brian Conway <[email protected]> wrote:
Are you using pf queues? I most frequently see that happen when there's no
space left in a queue. `pfctl -v -s queue`
This is the most likely cause.

Other than that, very busy unbound servers may also need
net.inet.udp.sendspace to be raised.


you might also want to check the following in unbound.conf on busy servers
num-threads
outgoing-range
num-queries-per-thread

G
ps. sorry Stuart for wrong reply

Reply via email to