Re: [Netconf] [NETCONF] <get-schema> mandatory parms discussion

[Andy Bierman <andy at netconfcentral.com>]

Mark Scott wrote:
> Martin and Andy,
> 
> Agreed we can make <format> and <version> optional so long as the server
> returns an error if more than one entry exists with the same
> <identifier>.


IMO, if <version> is empty, then the server needs to pick a version.
It SHOULD be the most recent version (perhaps MUST).

The module itself will indicate the revision, if the client
really cares.

The operator may not know the revision, or really care.

I prefer text that says:

  If the version parameter is missing, and the /netconf-state/schemas/schema
  list contains more than one entry for a specific identifier
  value, then the server SHOULD return the more recent version.

This is better for servers that will just have the most
recent version of accessible modules, and rarely have more
than 1 version of a groupings/typedefs module.


> 
> The revised <get-schema> negative responses (sec 3.1) proposed:
> 	If requested schema does not exist, the <error-tag> is
> 'invalid-value'.
> 	If requested schema is not unique, the <error-tag> is
> 'operation-failed' and the <error-app-tag> is 'data-not-unique'.
> 
> Default <format> of YANG is also specified in sec 3.1, and example 2b is
> updated to show default of YANG when format not specified.
> 
> and the revised RPC definition proposed:
> 	rpc get-schema {
> 	    description
>       	  "When the schema is available on the device this operation is
> 	         used to return it via NETCONF.  If requested schema
> does not exist,
>                the <error-tag> is 'invalid-value'.  If requested schema
> is not unique,
>                the <error-tag> is 'operation-failed' and the
> <error-app-tag> is
> 		   'data-not-unique'.";
> 	    input {
> 	      leaf identifier {
>       	  type string;
> 	        mandatory true;
> 	      }
>       	leaf version {
> 	        type string;
> 	      }
>       	leaf format {
> 	        type identityref {
>       	    base schema-format;
> 	        }
> 	      }
> 	    }
> 	    output {
> 	      anyxml data {
> 	        description "Contains the schema content.";
> 	      }
> 	    }
> 
> These updates will be made in v-09 update.  Any comments or concerns
> please raise them.  
> 
> cheers,
> Mark
> 


Andy

> -----Original Message-----
> From: Martin Bjorklund [mailto:mbj at tail-f.com] 
> Sent: Friday, October 02, 2009 6:23 AM
> To: mehmet.ersue at nsn.com
> Cc: andy at netconfcentral.com; Scott, Mark (CAR:2N00); netconf at ietf.org
> Subject: Re: [Netconf] [NETCONF] <get-schema> mandatory parms discussion
> 
> "Ersue, Mehmet (NSN - DE/Munich)" <mehmet.ersue at nsn.com> wrote:
>> Hi Mark, Andy, 
>>
>> modifying the the mandatory parameters in <get-schema> 
>> has been proposed earlier in September. There were no 
>> arguments against it at that time.
>>
>> In the sake of progress I would like to ask you to try 
>> to converge on a solution.
>> OTHERS: PLEASE speak up if you agree or disagree with 
>> any points made by either Mark or Andy.
> 
> I agree with Andy, i.e. <format> and <version> should be optional.  I
> think Mark's concern that the client might get a schema it didn't
> intended is covered by saying that if there are more than one schema
> with the given <identifier>, an error is returned.  In this case the
> client must give all parameters.
> 
> 
> /martin
>