BIND 10 #3036: Modify DHCP6 to generate NameChangeRequest

[BIND 10 Development]

#3036: Modify DHCP6 to generate NameChangeRequest
-------------------------------------+-------------------------------------
            Reporter:  marcin        |                        Owner:  tomek
                Type:  enhancement   |                       Status:
            Priority:  medium        |  reviewing
           Component:  dhcp-ddns     |                    Milestone:
            Keywords:                |  Sprint-DHCP-20130821
           Sensitive:  0             |                   Resolution:
         Sub-Project:  DHCP          |                 CVSS Scoring:
Estimated Difficulty:  0             |              Defect Severity:  N/A
         Total Hours:  0             |  Feature Depending on Ticket:
                                     |          Add Hours to Ticket:  0
                                     |                    Internal?:  0
-------------------------------------+-------------------------------------
Changes (by marcin):

 * owner:  marcin => tomek


Comment:

 Replying to [comment:11 tomek]:
 > > > '''ncr_msg.cc|h'''
 > > >
 > > > fromDUID(): Changes look ok, but there's one thing that requires a
 bit
 > > > though (or at least extra comment in the text). RFC4701, Section 3.5
 > > > says that FQDN should be in canonical format, as defined in RFC4034,
 > > > Section 6.2. That section talks about converting FQDN to lowercase.
 Do
 > > > we do that? If we do, we should make a brief comment about it and
 > > > point to specific place where this conversion happens. If we don't,
 we
 > > > should extend the code. I'm not saying that this is the best place.
 It
 > > > isn't. Probably somewhere in dns/labelsequence would be better (or
 in
 > > > places that use that class).
 > >
 > > That is a good point. The isc::dns::Name class is capable of doing it
 but by default it is disabled. I made a change in the DHCPv6 srv code to
 force downcase.
 > Are there any unit-tests that check that?

 I now leverage existing unit tests to cover that.

 > > I don't want to bloat this ticket anymore.
 > Then please create a separate ticket for that (or at least add a @todo
 in the test code).

 Created #3101.

 > > > Finally, the code doesn't build on Ubuntu 13.04 x64.
 > > Hopefully, it compiles now.
 > Nope. Still failed compilation. But the fix was easy. I pushed my
 changes. It compiles now.
 > Please pull.

 Thanks for this!

 >
 > > I don't understand the point about too long entry. Anything shorter
 would be vague and would not reflect what the new code is actually doing.
 Plus, this is not the final implementation and I don't want to make
 impression that we are doing any DNS updates at this time. So the last
 sentence is required.
 > Again, it was only a comment. Let it keep that way.
 >
 > I have one more minor comment, which probably doesn't require any code
 changes, just a bit of explanation.
 > Option6ClientFqdn::packDomainName() creates labels object from
 dereferenced domain_name_. What if the name is empty?
 >

 That's a great point.

 Actually, it requires changes to the code. I now check the domain_name
 pointer and if it is NULL, I do nothing.

-- 
Ticket URL: <http://bind10.isc.org/ticket/3036#comment:12>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development