BIND 10 #816: TSIG: update XFROUT to use TSIG

[BIND 10 Development]

#816: TSIG: update XFROUT to use TSIG
-------------------------------------+-------------------------------------
                   Reporter:         |                 Owner:  stephen
  stephen                            |                Status:  reviewing
                       Type:         |             Milestone:
  enhancement                        |  Sprint-20110614
                   Priority:  minor  |            Resolution:
                  Component:         |             Sensitive:  0
  xfrout                             |           Sub-Project:  DNS
                   Keywords:         |  Estimated Difficulty:  5.0
            Defect Severity:  N/A    |           Total Hours:  0
Feature Depending on Ticket:  tsig   |
        Add Hours to Ticket:  0      |
                  Internal?:  0      |
-------------------------------------+-------------------------------------
Changes (by zzchen_pku):

 * owner:  zzchen_pku => stephen


Comment:

 Replying to [comment:7 stephen]:
 > '''src/bin/xfrout/xfrout.py.in'''
 > dns_xfrout_start()
 > the check on rcode_ returned from _parse_query_message will not cope
 with errors other than FORMERR or NOTAUTH.  The logic should be:
 > {{{
 > if rcode_ == Rcode.NOTAUTH():
 >     return self._reply_query_with_error_rcode(msg, sock_fd, rcode_)
 > elif rcode_ != Rcode.NOERROR():
 >     return self._reply_query_with_format_error(msg, sock_fd)
 > }}}
 Good suggestion, updated.
 > _reply_xfrout_query()
 > Within _send_message_with_last_soa(), the length of the TSIG RR is added
 to the message_upper_len argument when checking against
 XFROUT_MAX_MESSAGE_SIZE.  However, if the main loop ends when the next
 packet to be sent should include a signature, on exit from the loop,
 message_upper_len will include the length of the TSIG RR.  In this case,
 the TSIG RR length is being included in the calculation twice.
 Yeah, I have update the handling logic.
 > set_tsig_key_ring()
 > Is the comment for this method correct?  It appears to apply to setting
 the TSIG key for a zone, not populating the key ring.
 Update the comment, and add a TODO item, which will be done in #943
 >
 > '''src/bin/xfrout/tests/xfrout_test.py.in'''
 > create_request_data_with_tsig()
 > There appears to be some line of code (starting
 'self.create_mock_tsig_ctx...') temporarily commented out.
 Removed it.
 > The "magic number" 655020 seems to be related to the variable
 XFROUT_MAX_MESSAGE_SIZE in the main code.  As it is the relationship
 between the two - rather than the absolute value - that appears to be
 important, I suggest it be defined as a variable equal to
 (XFROUT_MAX_MESSAGE_SIZE - 15) (together with a comment explaining why
 there is the relationship).
 Done.
 Please check , thanks for your review.

-- 
Ticket URL: <http://bind10.isc.org/ticket/816#comment:8>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development