[Freeipa-devel] [PATCH] 103 Fix automountkey-mod

[Rob Crittenden]

Martin Kosek wrote:
> On Thu, 2011-07-28 at 11:49 +0200, Martin Kosek wrote:
>> On Wed, 2011-07-27 at 14:53 -0400, Rob Crittenden wrote:
>>> Martin Kosek wrote:
>>>> On Wed, 2011-07-27 at 12:11 -0400, Dmitri Pal wrote:
>>>>> On 07/27/2011 12:00 PM, Martin Kosek wrote:
>>>>>> On Wed, 2011-07-27 at 10:41 -0400, Rob Crittenden wrote:
>>>>>>> Martin Kosek wrote:
>>>>>>>> Fix automountkey-mod so that automountkey attribute is correctly
>>>>>>>> updated. Add this test case to the unit tests.
>>>>>>>>
>>>>>>>> https://fedorahosted.org/freeipa/ticket/1528
>>>>>>> It fixes the problem but I've found another: --key isn't required so if
>>>>>>> you don't pass it in then a backtrace will occur:
>>>>>>>
>>>>>>> Traceback (most recent call last):
>>>>>>>      File "/home/rcrit/redhat/freeipa-master/ipaserver/rpcserver.py", line
>>>>>>> 220, in wsgi_execute
>>>>>>>        result = self.Command[name](*args, **options)
>>>>>>>      File "/home/rcrit/redhat/freeipa-master/ipalib/frontend.py", line
>>>>>>> 425, in __call__
>>>>>>>        ret = self.run(*args, **options)
>>>>>>>      File "/home/rcrit/redhat/freeipa-master/ipalib/frontend.py", line
>>>>>>> 731, in run
>>>>>>>        return self.execute(*args, **options)
>>>>>>>      File "/home/rcrit/redhat/freeipa-master/ipalib/plugins/automount.py",
>>>>>>> line 873, in execute
>>>>>>>        keys += (self.obj.get_pk(options['automountkey'],
>>>>>>> KeyError: 'automountkey'
>>>>>>>
>>>>>>> Also, automountinformation is already required. This may be a leftover
>>>>>>> from when we used it in description, this can probably be lifted too.
>>>>>>>
>>>>>>> rob
>>>>>> Good catch. I fixed this bug too and I also made --newinfo optional so
>>>>>> that automountkey may be just renamed without changing its info
>>>>>> attribute.
>>>>>>
>>>>>> I didn't bump up API VERSION as these are either compatible changes or
>>>>>> they caused server internal error.
>>>>>>
>>>>>> Martin
>>>>>
>>>>> Should the ticket be moved into 2.1 July sprint then?
>>>>
>>>> Yes, I would like this to be included in 2.1. I will move ticket to
>>>> correct milestone (2.1) if we manage to review&push it before release.
>>>>
>>>> Martin
>>>
>>> nack. Something is up with _mod. I can't be sure it is this patch or it
>>> was always here.
>>>
>>> In the UI every change wanted to try to rename the entry. On the
>>> command-line I wasn't able to update the info at all.
>>>
>>> rob
>>
>> Hm, I think this problem was in the _mod command all the time.
>> 'description' field was being filled every time which triggered rename
>> operation. This caused problems.
>>
>> I rewrote _mod command so that 'description' (i.e. rename) is filled
>> only when needed.
>>
>> I checked UI and automountkey_mod command worked OK for me.
>>
>> Martin
>
> I revisited automountkey_mod command as there was another corner case.
> get_pk() was changed to include automountinfo in RDN only for direct
> maps.
>
> All tests are OK, even Rob's extended automount test preview. Web UI
> behaved ok too.
>
> Martin

ack, pushed to master and ipa-2-1