BIND 10 #2421: don't reject configuring entire inmemory datasrc due to one broken zone

[BIND 10 Development]

#2421: don't reject configuring entire inmemory datasrc due to one broken zone
-------------------------------------+-------------------------------------
                   Reporter:         |                 Owner:  muks
  jinmei                             |                Status:  reviewing
                       Type:         |             Milestone:
  defect                             |  Sprint-20121120
                   Priority:         |            Resolution:
  medium                             |             Sensitive:  0
                  Component:  data   |           Sub-Project:  DNS
  source                             |  Estimated Difficulty:  4
                   Keywords:         |           Total Hours:  0
            Defect Severity:  High   |
Feature Depending on Ticket:         |
        Add Hours to Ticket:  0      |
                  Internal?:  0      |
-------------------------------------+-------------------------------------

Comment (by muks):

 Hi Michal, Jinmei

 Replying to [comment:9 vorner]:
 > I agree with Jinmei it would be better to have some kind of exception
 hierarchy for this.

 This is added now.

 > Also, there are some problems with the code:
 >  * You reuse log-ids. That's forbidden, each log ID must appear at
 exactly one place of the code.

 Fixed. :)

 >  * Can a user cause `DATASRC_LOAD_NULL_RRSET` to happen? How? I think
 this is a programmer error (that there's a NULL RRset to start with). What
 would the user do with a null RRset if it happens? Would he at least guess
 what the problem is? (because from user point of view, there's no such
 thing as NULL RRset).

 There was a way before using the `add()` method which has been removed
 since. But even then, I feel it's better to keep the validation check in
 place. But we don't catch the NullRRset (it's not in the `ZoneException`
 hierarchy, but is an `InvalidParameter`) anymore.

 >  * Could there also be unit tests for this, besides lettuce tests?

 I feel lettuce is the best place for it as this is a bug that shows up
 with bad user configuration. One more check has been added to the lettuce
 test. I think they cover the various cases now. Please see this and if you
 feel that a unit test would cover something and would be needed, I'll add
 it.

-- 
Ticket URL: <http://bind10.isc.org/ticket/2421#comment:10>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development