[Yeti DNS Discuss] Strange BIND resolver behaviour (may be not Yeti-specific)
Stephane Bortzmeyer
bortzmeyer at nic.fr
Tue Jul 14 20:59:00 UTC 2015
I take the opportunity of Yeti work to observe the behavior of my
resolvers, something that I don't typically do for the "normal"
root. So, this behaviour may be not Yeti-specific.
Tcpdumping the traffic of a BIND 9.9.5 resolver, I see many requests
which, IMHO, should have been answered from cache:
22:13:14.088947 IP6 2a01:e35:8bd9:8bb0:21e:8cff:fe76:29b6.46828 > 2001:559:8000::6.53: 42576% [1au] DS? 192.in-addr.arpa. (45)
22:13:14.261736 IP6 2001:559:8000::6.53 > 2a01:e35:8bd9:8bb0:21e:8cff:fe76:29b6.46828: 42576- 0/10/13 (729)
22:13:18.487354 IP6 2a01:e35:8bd9:8bb0:21e:8cff:fe76:29b6.59531 > 2a02:2810:0:405::250.53: 18213% [1au] DS? 204.in-addr.arpa. (45)
22:13:18.542828 IP6 2a02:2810:0:405::250.53 > 2a01:e35:8bd9:8bb0:21e:8cff:fe76:29b6.59531: 18213- 0/13/9 (832)
Why the 204.in-addr.arpa request, four seconds after a successful
one for 192.in-addr.arpa? BIND certainly should have cached the
delegation for .arpa.
22:39:36.840415 IP6 2a01:e35:8bd9:8bb0:21e:8cff:fe76:29b6.47604 > 2001:559:8000::6.53: 49133% [1au] DS? twitter.com. (40)
22:39:37.009472 IP6 2001:559:8000::6.53 > 2a01:e35:8bd9:8bb0:21e:8cff:fe76:29b6.47604: 49133- 0/15/16 (735)
22:40:39.667999 IP6 2a01:e35:8bd9:8bb0:21e:8cff:fe76:29b6.48576 > 2a02:cdc5:9715:0:185:5:203:53.53: 60622% [1au] DS? twimg.com. (38)
22:40:39.724273 IP6 2a02:cdc5:9715:0:185:5:203:53.53 > 2a01:e35:8bd9:8bb0:21e:8cff:fe76:29b6.48576: 60622- 0/15/16 (733)
Same thing here, the delegation for .com should have been in the
cache.
Any opinion?
More information about the discuss
mailing list