BIND 10 #1378: Configuration of #213

BIND 10 Development do-not-reply at isc.org
Tue Nov 8 17:48:05 UTC 2011


#1378: Configuration of #213
-------------------------------------+-------------------------------------
                   Reporter:         |                 Owner:  jinmei
  vorner                             |                Status:  reviewing
                       Type:  task   |             Milestone:
                   Priority:  major  |  Sprint-20111122
                  Component:  Boss   |            Resolution:
  of BIND                            |             Sensitive:  0
                   Keywords:         |           Sub-Project:  Core
            Defect Severity:  N/A    |  Estimated Difficulty:  0
Feature Depending on Ticket:         |           Total Hours:  0
        Add Hours to Ticket:  0      |
                  Internal?:  0      |
-------------------------------------+-------------------------------------

Comment (by jinmei):

 Overall it looks fine.  I have some relatively minor comments and
 questions.

 '''bind10_src.py.in'''
 - read_bind10_config and start_all_processes are now both sufficiently
   small and could be combined into a single method.  The only reason
   we don't do this seems to be the convenience for tests, and, if
   that's the reason I'd explicitly add a note to read_bind10_config
   (otherwise someone would try "refactoring" the code later).  Also,
   if that's the reason we might want to make read_bind10_config
   "protected" by renaming it to `_read_bind10_config`.
 - start_all_processes: not really for this branch, but I'd make this
   method a bit more concise by removing mostly redundant temporary
   variable:
 {{{
         #c_channel_env = self.c_channel_env => remove this line
         self.start_ccsession(self.c_channel_env)
 }}}

 '''unit tests'''
 - MockBob.start_simple(): why this change?
 {{{
 -                    'b10-xfrin': self.start_xfrin,
 }}}
   (why was xfrin removed while keeping others?)
 - bind10_test: I don't think removing
   test_start(none/auth/resolver/both) is really correct (maybe except
   for none), because in test_config_start we only check the code path
   from config_handler while test_start variants check the one from
   start_all_processes (and testing these cases shouldn't be difficult
   anyway).
 - test_config_start_once: what's the purpose of this addition?
 {{{
 +        bob._BoB_started = True
 }}}
   (same for test_start_dhcp, etc)

 - TestBossComponents: is there anything that has changed from the
   original trac213 tests?  (I'm asking whether I need to review this
   again)

 '''system tests'''
 I guess this hack is due to a bug already fixed in master (forgot the
 ticket number):
 {{{
 +echo 'config add Boss/components x
 +config remove Boss/components x
 }}}
 I'd rather merge that fix here and remove this kludge now.

-- 
Ticket URL: <http://bind10.isc.org/ticket/1378#comment:5>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development


More information about the bind10-tickets mailing list