Ok please file a ticket and we can investigate.

Thanks!



On Tue, Sep 9, 2014 at 12:54 PM, Romain Richard <
[email protected]> wrote:

> Here's a summary:
>
>                 | with --diff | without --diff
> ----------------------------------------------
> with --check    | changes     | changes
> ----------------------------------------------
> without --check | no changes  | no changes
>
>
> On Tuesday, September 9, 2014 9:42:05 AM UTC-7, Michael DeHaan wrote:
>>
>> I'm wondering if this may be because it doesn't have permission to read
>> them and the --diff flag has a buglet in it.
>>
>> check should be generally fine, I would suspect the diff logic could be
>> throwing it.
>>
>> Does it report a change w/o --diff ?
>>
>>
>>
>> On Tue, Sep 9, 2014 at 12:32 PM, Romain Richard <romain.richar...@
>> gmail.com> wrote:
>>
>>> Installed Ansible 1.7.1 from the ppa:
>>>
>>> $ ansible --version
>>> ansible 1.7.1
>>>
>>> And I am seeing the exact same output as with the 1.6.3, the --check
>>> still reports the the authorized_keys files on the remote host are empty
>>> (the diff is still the same).
>>>
>>> On Tuesday, September 9, 2014 8:47:06 AM UTC-7, Michael DeHaan wrote:
>>>>
>>>> Ansible 1.6.3 is no longer the active released version of Ansible, and
>>>> since 1.6.3 there have been many updates, many security related.
>>>>
>>>> When reporting issues, it's helpful to have tested at least the latest
>>>> release, which is 1.7.1.
>>>>
>>>> If you see diff issues there, let us know, but seeing you reported on
>>>> 1.6.3 there's a good chance this is now resolved.
>>>>
>>>> Thanks!
>>>>
>>>> On Mon, Sep 8, 2014 at 7:26 PM, Romain Richard <romain.richar...@
>>>> gmail.com> wrote:
>>>>
>>>>> Thanks for your reply, more info below.
>>>>>
>>>>> On Monday, September 8, 2014 4:01:30 PM UTC-7, Michael DeHaan wrote:
>>>>>>
>>>>>> This could be handled by having a previous task that copied a blank
>>>>>> file over, provided you weren't logged in as that user.
>>>>>>
>>>>>
>>>>> That's not a bad idea, I will look into that.
>>>>>
>>>>>
>>>>>> You could also keep a list of previous keys and use state=absent to
>>>>>> remove those.
>>>>>>
>>>>>
>>>>> That seems cumbersome.
>>>>>
>>>>>
>>>>>> I'm open to the idea of having a parameter like exclusive=yes that
>>>>>> removes the other keys in the file.
>>>>>>
>>>>>
>>>>> Would sure make my task easier.
>>>>>
>>>>>
>>>>>> Some ansible modules don't fully understand check mode and will
>>>>>> report "changed=True" automatically without running in check mode rather
>>>>>> than risk making a change.
>>>>>>
>>>>>
>>>>> I see.
>>>>>
>>>>>
>>>>>> Can we see the changed lines from your ansible playbook, as well as
>>>>>> the output of ansible --version to confirm this is from those lines and a
>>>>>> recent version of Ansible?
>>>>>>
>>>>>
>>>>> $ ansible --version
>>>>> ansible 1.6.3
>>>>>
>>>>> Not sure what you meant by "the changed lines from your ansible
>>>>> playbook".
>>>>>
>>>>> In this case it is showing that there would be additions from your
>>>>>> template that are not in the original file, so it seems that it is
>>>>>> returning accurately in this regard.
>>>>>> Or is your assertion that the diff is *also* wrong?  That seems
>>>>>> somewhat unlikely, but somewhat resembles what may be an older bug in
>>>>>> Ansible -- I could be wrong.
>>>>>>
>>>>>
>>>>> It seems that the diff is saying the same thing as the check, so I
>>>>> suppose it is not wrong, but it shows differences while there are actually
>>>>> none (if I had run the command again without the --check, there would have
>>>>> been no changes).
>>>>> It makes me believe that what Ansible feeds to the diff is wrong,
>>>>> because of that --check option.
>>>>>
>>>>> --
>>>>> You received this message because you are subscribed to the Google
>>>>> Groups "Ansible Project" group.
>>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>>> an email to [email protected].
>>>>> To post to this group, send email to [email protected].
>>>>> To view this discussion on the web visit https://groups.google.com/d/
>>>>> msgid/ansible-project/0145f866-b16c-4728-9ffa-483d2bf9e451%
>>>>> 40googlegroups.com
>>>>> <https://groups.google.com/d/msgid/ansible-project/0145f866-b16c-4728-9ffa-483d2bf9e451%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>> .
>>>>>
>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>
>>>>
>>>>  --
>>> You received this message because you are subscribed to the Google
>>> Groups "Ansible Project" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to [email protected].
>>> To post to this group, send email to [email protected].
>>> To view this discussion on the web visit https://groups.google.com/d/
>>> msgid/ansible-project/6622c2e9-61d9-4af8-bf4a-
>>> ae40be65cb27%40googlegroups.com
>>> <https://groups.google.com/d/msgid/ansible-project/6622c2e9-61d9-4af8-bf4a-ae40be65cb27%40googlegroups.com?utm_medium=email&utm_source=footer>
>>> .
>>>
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>
>>  --
> You received this message because you are subscribed to the Google Groups
> "Ansible Project" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To post to this group, send email to [email protected].
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/ansible-project/5d37690c-90ac-40ae-be1c-0830e3970e27%40googlegroups.com
> <https://groups.google.com/d/msgid/ansible-project/5d37690c-90ac-40ae-be1c-0830e3970e27%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"Ansible Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/ansible-project/CA%2BnsWgzZgJu6rWaN3HG1tOViCBNB%2BNjSOiEGpjmGx_p8wRuVqg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to