Re: [yang-doctors] when statement of an optional leaf? Does it make an optional leaf?

["Jean Quilbeuf -X (jquilbeu - LIANEO at Cisco)" <[e-mail]>]

Hi Benoît, all,
To clarify:
by "it works" I meant that instances that do not specify a "maintenance-contact" nor a "under-maintenance" (which then default to false) become valid for ConfD after the modification. Did not try pyang !

Best,
Jean



-----Original Message-----
From: Benoit Claise <[e-mail]> 
Sent: Tuesday, March 17, 2020 7:08 PM
To: Ladislav Lhotka <[e-mail]>; Benoit Claise <bclaise=[e-mail]>; YANG Doctors <[e-mail]>
Cc: Eric Vyncke (evyncke) <[e-mail]>; Jean Quilbeuf -X (jquilbeu - LIANEO at Cisco) <[e-mail]>
Subject: Re: [yang-doctors] when statement of an optional leaf? Does it make an optional leaf?

Hi Lada,

Thanks for the XPath correction.
When I apply this change, I still see the maintenance-contact as mandatory.

$ pyang -f tree ietf-service-assurance.yang
module: ietf-service-assurance
   +--ro assurance-graph-version?       yang:counter32
   +--ro assurance-graph-last-change?   yang:date-and-time
   +--rw subservices
      +--rw subservice* [type id]
         +--rw type                                identityref
         +--rw id                                  string
         +--ro last-change? yang:date-and-time
         +--ro label?                              string
         +--rw under-maintenance?                  boolean
         +--rw maintenance-contact                 string

I guess the validation is not clever enough to combine "mandatory true" 
and Xpath (when "../under-maintenance='true'") from maintenance-contact with the optional under-maintenance leaf ... to conclude that maintenance-contact is actually optional.

Regards, Benoit
> Hi Benoit,
>
> "maintenance-contact" may remain mandatory - due to its default value, "under-maintenace" is always present, as far as XPath evaluation is concerned.
>
> But because of this, your when expression probably doesn't have the 
> effect that you expect: it will always be true. I think it needs to be 
> changed to
>
>      when "../under-maintenance = 'true'";
>
> Lada
>
> Benoit Claise <bclaise=[e-mail]> writes:
>
>> YANG doctors,
>>
>> $ pyang -f tree  ietf-service-assurance.yang
>> module: ietf-service-assurance
>>     +--ro assurance-graph-version?       yang:counter32
>>     +--ro assurance-graph-last-change?   yang:date-and-time
>>     +--rw subservices
>>        +--rw subservice* [type id]
>>           +--rw type                                identityref
>>           +--rw id                                  string
>>           +--ro last-change?                        yang:date-and-time
>>           +--ro label?                              string
>>           +--rw under-maintenance?                  boolean
>>           +--rw maintenance-contact string <===============================
>>           +--rw (parameter)?
>>           |  +--:(service-instance-parameter)
>>           |     +--rw service-instance-parameter
>>           |        +--rw service?         string
>>           |        +--rw instance-name?   string
>>           +--ro health-score?                       uint8
>>           +--rw symptoms
>>           |  +--ro symptom* [start-date-time id]
>>           |     +--ro id                     string
>>           |     +--ro health-score-weight?   uint8
>>           |     +--ro label?                 string
>>           |     +--ro start-date-time        yang:date-and-time
>>           |     +--ro stop-date-time?        yang:date-and-time
>>           +--rw dependencies
>>              +--rw dependency* [type id]
>>                 +--rw type               -> /subservices/subservice/type
>>                 +--rw id                 -> 
>> /subservices/subservice[type=current()/../type]/id
>>                 +--rw dependency-type?   identityref
>>
>> The YANG modules comes from
>> https://datatracker.ietf.org/doc/draft-claise-opsawg-service-assuranc
>> e-yang/
>> Have a look at
>> https://raw.githubusercontent.com/YangModels/yang/ecb622b214e59c5f631
>> 2e915d4d65823a6485852/experimental/ietf-extracted-YANG-modules/ietf-s
>> [e-mail]
>>
>> The leafs in question:
>>
>>             leaf under-maintenance {
>>               type boolean;
>>               default false;
>>               description
>>                 "An optional flag indicating whether this particular subservice is under
>>                 maintenance. Under this circumstance, the subservice symptoms and the
>>                 symptoms of its dependencies in the assurance graph should not be taken
>>                 into account. Instead, the subservice should send a 'Under Maintenance'
>>                 single symptom.
>>
>>                 The operator changing the under-maintenance value must set the
>>                 maintenance-contact variable.
>>
>>                 When the subservice is not under maintenance any longer, the
>>                 under-maintenance flag must return to its default value and
>>                 the under-maintenance-owner variable deleted.";
>>             }
>>             leaf maintenance-contact {
>>      when "../under-maintenance";
>>               type string;
>>               mandatory true;
>>               description
>>                 "A string used to model an administratively assigned name of the
>>                 resource that changed the under-maintenance value to 'true.
>>
>>                 It is suggested that this name contain one or more of the following:
>>                 IP address, management station name, network manager's name, location,
>>                 or phone number. In some cases the agent itself will be the owner of
>>                 an entry. In these cases, this string shall be set to a string
>>                 starting with 'monitor'.";
>>             }
>>
>> In case of a leaf (maintenance-contact) with a when statement 
>> pointing to an optional leaf (under-maintenance), should the leaf
>> (maintenance-contact) be marked as optional in the tree output?
>>
>> Regards, Benoit
>> _______________________________________________
>> yang-doctors mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/yang-doctors