[bind10-dev] C++ TSIG API: design/code/next steps
JINMEI Tatuya / 神明達哉
jinmei at isc.org
Wed Apr 27 08:40:48 UTC 2011
I've completed an initial step of the whole TSIG related tasks:
C++ API (with implementation) for TSIG signing. It's in ticket #812
and the implementation is in the review queue. Since many other
things explicitly or implicitly rely on it, I'm sharing some
information about it.
The basic design (that also covers the verify scenario to some extent)
is documented in the source code. Here's a copy of an HTML version of
the doxygen document:
http://bind10.isc.org/~jinmei/bind10/cpp/classisc_1_1dns_1_1_t_s_i_g_context.html
http://bind10.isc.org/~jinmei/bind10/cpp/classisc_1_1dns_1_1_t_s_i_g_record.html
(the design will be reviewed as part of code review, but if someone in
this list has comments on it, that would also be very helpful)
If we want to parallelize the remaining work, we may be able to do
something like this, once we agree on the basic API design:
- when we agree on the design and basic interface of the classes, we
can start the verify part (#813) based on the latest branch of
trac812. If no one else is interested in it I'm willing to take on
it, but someone is interested in this work it may be better in that
we can have a second eye on the design.
- likewise, when we agree on the design and basic interface, we can
probably start implementing the python wrapper (#814) for the
signing part. (With this approach we'd divide the task into
sign/verify parts, too)
- likewise, when we agree on the design and basic interface, we can
start #856 (using TSIG in our experimental "host" implementation).
This is probably a 30min job, and would be better to be done by
someone other than me so that we can see how easy/difficult the API
is to use.
- If the python wrapper for signing is implemented, we could also
start developing xfrin (it cannot verify the response, but at least
it can send AXFR requests with a valid TSIG sign).
---
JINMEI, Tatuya
More information about the bind10-dev
mailing list