Re: [nfsv4] Mirja Kühlewind's No Objection on draft-ietf-nfsv4-rpcrdma-cm-pvt-data-07: (with COMMENT)

[Mirja Kuehlewind <[e-mail]>]

Hi Chuck,

See below.

> On 19. Feb 2020, at 21:46, Chuck Lever <[e-mail]> wrote:
> 
> Hello Mirja-
> 
>> On Feb 19, 2020, at 9:24 AM, Mirja Kühlewind via Datatracker <[e-mail]> wrote:
>> 
>> Mirja Kühlewind has entered the following ballot position for
>> draft-ietf-nfsv4-rpcrdma-cm-pvt-data-07: No Objection
>> 
>> When responding, please keep the subject line intact and reply to all
>> email addresses included in the To and CC lines. (Feel free to cut this
>> introductory paragraph, however.)
>> 
>> 
>> Please refer to https://www.ietf.org/iesg/statement/discuss-criteria.html
>> for more information about IESG DISCUSS and COMMENT positions.
>> 
>> 
>> The document, along with other ballot positions, can be found here:
>> https://datatracker.ietf.org/doc/draft-ietf-nfsv4-rpcrdma-cm-pvt-data/
>> 
>> 
>> 
>> ----------------------------------------------------------------------
>> COMMENT:
>> ----------------------------------------------------------------------
>> 
>> One quick thought/comment: Another option for extensibility would be to use one
>> of the reserved flags to e.g. extend the fields of the private data field.
>> However, the draft states at all reserved flags need to be zero with version 1.
>> This seems to be a bit of a waste of space but moreover it's a lost opportunity
>> for an easy way to extend the private data field. Why was that decided?
> 
> The common way to extend this kind of data structures, I assumed,
> was to use the version field. The set of defined flag fields cannot
> change without a bump in the version number.

Yes that is a good and common way. However, usually if you have reserved flags, there is also a way to make use of those flags without changing the version which seem not the case here.

> 
> I would be interested to consider another way to handle extension,
> though at this point there is already an implementation of this
> protocol, which limits our ability to adjust the structure format
> too much.

I would not think that this would change anything in the current implementation, as the current implementation should just ignore the flags. However, this is no big deal, but maybe it’s something to think about if there is a way to make better use of these reserved flags/bits. 

Mirja


> 
> Thanks for your comment!
> 
> 
> --
> Chuck Lever
> 
> 
> 
>