Discussion:
how to verify Pipelined TCP and DNSCOOKIE
Rodrigo Reyna
2018-09-04 13:46:16 UTC
Permalink
Hi,there.

Please someone help me.
I am currently verifying the function of BIND 9.11, but I am in two
troubles.

1.
I don't know how to confirm [Pipelined TCP queries(RFC7766)].
Specifically, please tell me if there is a method to check with the command
such as dig for the server on which BIND 9.11 is running.

2.
The other is about [DNSCOOKIE].
Querying servers with DNSCOOKIE enabled / disabled with dig (bind 9.
11.4),it is always displayed as DNSCOOKIE: xxxxx (good).

Why does it appear as good even if I query the server with DNSCOOKIE
disabled?
Also,please tell me if you know a command that will cause DNSCOOKIE to fail
(ie, other than good is displayed).

Thank you.



--
Sent from: http://bind-users-forum.2342410.n4.nabble.com/
_______________________________________________
Please visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe from this list

bind-users mailing list
bind-***@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users
Rodrigo Reyna
2018-09-04 13:55:24 UTC
Permalink
Hi,there.

Please someone help me.
I am currently verifying the function of BIND 9.11, but I am in two
troubles.

1.
I don't know how to confirm [Pipelined TCP queries(RFC7766)].
Specifically, please tell me if there is a method to check with the command
such as dig for the server on which BIND 9.11 is running.

2.
The other is about [DNSCOOKIE].
Querying servers with DNSCOOKIE enabled / disabled with dig (bind 9.
11.4),it is always displayed as DNSCOOKIE: xxxxx (good).

Why does it appear as good even if I query the server with DNSCOOKIE
disabled?
Also,please tell me if you know a command that will cause DNSCOOKIE to fail
(ie, other than good is displayed).

Thank you.



--
Sent from: http://bind-users-forum.2342410.n4.nabble.com/
_______________________________________________
Please visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe from this list

bind-users mailing list
bind-***@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users
Tony Finch
2018-09-04 14:13:00 UTC
Permalink
Post by Rodrigo Reyna
I don't know how to confirm [Pipelined TCP queries(RFC7766)].
Specifically, please tell me if there is a method to check with the command
such as dig for the server on which BIND 9.11 is running.
`mdig` comes with BIND and does multiple concurrent queries, so you can
check pipelining behaviour like this:

$ rndc flush; mdig +tcp +noall +answer @::1 \
www.melbourne.edu.au \
www.u-tokyo.ac.jp \
www.uonbi.ac.ke \
www.puc-rio.br \
web.mit.edu \
www.cam.ac.uk

A good selection of sites near and far should nicely demonstrate
out-of-order replies. If you don't flush the cache first then you'll get
the answers in an effectively random order.

For years I have used `adns` for bulk concurrent queries, and it's really
good at pipelining queries over a TCP connection.
Post by Rodrigo Reyna
The other is about [DNSCOOKIE].
Dunno this one I'm afraid.

Tony.
--
f.anthony.n.finch <***@dotat.at> http://dotat.at/
a fair, free and open society
_______________________________________________
Please visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe from this list

bind-users mailing list
bind-***@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users
Ray Bellis
2018-09-04 14:40:23 UTC
Permalink
Post by Tony Finch
`mdig` comes with BIND and does multiple concurrent queries, so you can
...
A good selection of sites near and far should nicely demonstrate
out-of-order replies. If you don't flush the cache first then you'll get
the answers in an effectively random order.
Technically, pipelining is not the same thing as "out of order replies",
although the former is required for the latter to work.

Pipelining just means that you can submit multiple queries over the same
socket without having to wait for any answers at all, and without the
server flushing its input queue.

AFAIK BIND was always capable of this - the new bit is that BIND will
consume the queries and start processing them (and send the responses
back in whatever order they become available) rather than processing
them strictly in order with each query and response proceeding in
lock-step mode.

This is a distinction that I still have to explain to my RFC 7766
co-authors on occasion :)

kind regards,

Ray
_______________________________________________
Please visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe from this list

bind-users mailing list
bind-***@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users

Loading...