Darren,
On 08/29/11 12:07, Darren Kenny wrote:
> Hi Matt,
>
> Sorry for not being clearer...
>
> On 29/08/2011 11:52, Matt Keenan wrote:
>> Hey Darren,
>>
>> thanks for the review, but not 100% sure what you mean here.
>>
>> On 08/29/11 10:56, Darren Kenny wrote:
>>> Hi Matt,
>>>
>>> I'm not sure about the location if the check_init_SMF() call... I would
>>> expect
>>> that this is only run when you call the python module directly, but not if
>>> done
>>> via NOSE.
>>>
>> Do you mean the location of the method definition or when it's actually
>> called ?
> I mean when it's called.
>
>> It should really be called regardless of whether running Nose or not,
>> the tests that call
>> check_install_SMF() will fail if the service is not there, and this is
>> regardless of whether
>> being run directly (manually) or via nose.
> Ok, but in the way you have the code in test_aimdns_manual.py, you have the
> only
> call to the method within a clause like:
>
> 248 if __name__ == '__main__':
> 249 # Ensure system/install/server SMF service is available
> 250 check_install_SMF()
> 251
> 252 gettext.install("ai", "/usr/lib/locale")
> 253 unittest.main()
>
> Any thing inside of this clause is only ever run if you call the code
> directly as:
>
> python test_aimdns_manual.py
>
> It will not be run if you call it via the NOSE test harness, which will import
> it, and as such the module name (__name__) will not be '__main__'.
>
Ah now I get it, I'll move check_install_SMF() to be outside of this if.
I was "manually"
running "python test/manual/test_aimdns_manual.py", and thus had not
seen that it
would not be run with Nose.
cheers
Matt
>>> As such it would seem wrong to be throwing a NOSE specific exception when
>>> not
>>> running NOSE...
>>>
>> Fair point, i suppose. what do you suggest, just exiting without raising
>> anything ?
> No, it's probably fine to raise it, I'm just trying to say that it will never
> be
> raised within NOSE as coded.
>
>>> Maybe it should be called outside of the 'if' statement, that way it will
>>> always
>>> be executed.
>>>
>> Outside of what if statement ? what source file are you referring to here.
> See above.
>
> Hope this is clearer.
>
> Thanks,
>
> Darren.
>
>> cheers
>>
>> Matt
>>
>>> Thanks,
>>>
>>> Darren.
>>>
>>> On 26/08/2011 18:13, Matt Keenan wrote:
>>>> Hi,
>>>>
>>>> Re-Worked webrev now posted to ;
>>>> https://cr.opensolaris.org/action/browse/caiman/mattman/7050547/
>>>>
>>>> This webrev contains :
>>>>
>>>> - Splitting out tests that require 120 second timeouts to be manual
>>>> tests, and leaving
>>>> all other tests intact.
>>>>
>>>> - Calling check_install_SMF() at init time where it makes sense.
>>>>
>>>> cheers
>>>>
>>>> Matt
>>>>
>>>> On 08/26/11 14:09, Matt Keenan wrote:
>>>>> Hi,
>>>>>
>>>>> Can I get a review and blessing for fix for :
>>>>> 7050547 ai_sd/aimdns unit tests failing on machine not configured
>>>>> as AI server
>>>>> http://monaco.sfbay.sun.com/detail.jsf?cr=7050547
>>>>>
>>>>> Webrev:
>>>>> https://cr.opensolaris.org/action/browse/caiman/mattman/7050547/
>>>>>
>>>>>
>>>>> Slim install unit tests tend to be run on a build machine which is not
>>>>> configured as an install server, thus these tests fail. Nose provides
>>>>> a SkipTest exception which when used will not report the tests
>>>>> themselves as failed.
>>>>>
>>>>> For ai_sd if multicast DNS SMF is not available then the tests will fail.
>>>>>
>>>>> For aimdns if install/server SMF is not available 5 tests will fail.
>>>>>
>>>>> I also discovered that when a service is removed, it remains
>>>>> advertised for a default time of 120 seconds, this results in
>>>>> time.sleep(120) calls being required for tests to be successful.
>>>>>
>>>>> cheers
>>>>>
>>>>> Matt
>>>>> _______________________________________________
>>>>> caiman-discuss mailing list
>>>>> [email protected]
>>>>> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
>>>> _______________________________________________
>>>> caiman-discuss mailing list
>>>> [email protected]
>>>> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
_______________________________________________
caiman-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
