Re: [nfsv4] Milestone draft for working group review

Chuck Lever <> Mon, 21 August 2017 14:26 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id B779E13202D for <>; Mon, 21 Aug 2017 07:26:47 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -4.22
X-Spam-Status: No, score=-4.22 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_MED=-2.3, RCVD_IN_MSPIKE_H4=-0.01, RCVD_IN_MSPIKE_WL=-0.01, SPF_PASS=-0.001, UNPARSEABLE_RELAY=0.001] autolearn=ham autolearn_force=no
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id co2uA8027f7c for <>; Mon, 21 Aug 2017 07:26:45 -0700 (PDT)
Received: from ( []) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by (Postfix) with ESMTPS id 8F78513247A for <>; Mon, 21 Aug 2017 07:26:45 -0700 (PDT)
Received: from ( []) by (Sentrion-MTA-4.3.2/Sentrion-MTA-4.3.2) with ESMTP id v7LEQidU022769 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 21 Aug 2017 14:26:44 GMT
Received: from ( []) by (8.14.4/8.14.4) with ESMTP id v7LEQhlS015245 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 21 Aug 2017 14:26:44 GMT
Received: from ( []) by (8.14.4/8.14.4) with ESMTP id v7LEQh3R011238; Mon, 21 Aug 2017 14:26:43 GMT
Received: from (/ by default (Oracle Beehive Gateway v4.0) with ESMTP ; Mon, 21 Aug 2017 07:26:43 -0700
Content-Type: text/plain; charset="utf-8"
Mime-Version: 1.0 (Mac OS X Mail 9.3 \(3124\))
From: Chuck Lever <>
In-Reply-To: <>
Date: Mon, 21 Aug 2017 10:26:42 -0400
Cc: "" <>
Content-Transfer-Encoding: quoted-printable
Message-Id: <>
References: <>
To: David Noveck <>
X-Mailer: Apple Mail (2.3124)
X-Source-IP: []
Archived-At: <>
Subject: Re: [nfsv4] Milestone draft for working group review
X-Mailman-Version: 2.1.22
Precedence: list
List-Id: NFSv4 Working Group <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Mon, 21 Aug 2017 14:26:48 -0000

Happy eclipse day, Dave -

> On Aug 17, 2017, at 7:35 AM, David Noveck <> wrote:
> We've previously discussed and reviewed a proposal for a new working group charter.  The only thing missing was a set of proposed milestones.
> Since that time, I've been discussing this subject with the authors of active working group dcuments and I-D's that appear likely to become working group documents (except for Benny HaLevy who seems to no longer be active in the working group).  The draft of what we (Chuck, Tom, Christoph and I) have come up with is attached, so that the working group can review it.  I had assumed that others in the working group did not have any proposed milestones that they would like to include and work on.  If I was wrong about that, please bring any proposed new milestones to the attention of the working group as part of this review.  If you have an idea for a milestone which is not quite ready for this list, hold that thought.  There is a mechanism for the working group to create new milestones later, once the charter is approved.
> It would be good if the working group could review this by Friday 8/25, so that we can make any necessary adjustments and provide the chairs and ads with a consolidated charter proposal including the stuff I sent to them on 7/31 but with the milestones included.  If anyone thinks theey need more time to do this review, please let me know.
> Most of the items on this list are not yet working group documents but I hope I haven't included anything that people would object to becoming working group documents.  If you do object or have issues with any of these becoming working group douments, we need to address that as part of the milestone review.
> There are some notes below about the documents associated with the proposed milestones, in time order.  Please let me know if you need any clarification.
> First are documents associated with the 2018 milestones:  
> The 3/2018 milestone is for the working group document deriving from draft-hellwig-nfsv4-scsi-layout-nvme.  This document was discussed at IETF99 and there was general agreement that it would become standards-track.  There were a number of structural changes proposed at that time and I assume Christoph is dealing with those before formally requesting working group status for this document.
> The 6/2018-10/2018 milestones all concern documents relating to changes for trunking and state migration.  I presented slides about these at IETF99.  If these are not easily available, please let me know.  There are four milestones and three documents as follows:
> 	• The 6/2018 milestone concerns draft-ietf-nfsv4-migration-issues.  It has become clear that fixing migration and trunking are inherently related so -13 has been retitled to reflect this fact.  If people have the time, I'd appreciate reviews of this document to make sure that we adress issue now, rather than in a rush at WGLC.  Although this document is informational, there will be a number of standards-track documents that derive from it.
> 	• The 8/2018 milestone concerns a document addressing trunking discovery in NFSv4.0.  This document addresses the same issues that had been dealt with in Andy's previous trunking discovery document although the treatment is significantly different.  In the IETF99 slides, the prospective I-D on which this will be based is referred to as draft-adamson-nfsv4-mv0-truunking-update.  
> Since then, Andy has retired and his co-author Chuck Lever has agreed to submit draft-cel--nfsv4-mv0-trunking-update, based on the work Andy has already done.  I think Chuck will work on submitting this as soon as he has had an opprtuunity to catch his breath after completing the work on the last of the documents associated with RPC-over-RDMA version 1.  I expect to be a co-author of the trunking update document and will work to keep the trunking discussion for both minor versions consistent.
> 	• The two 10/2018 milestones will most likely be satisfied by the same docment, specfically the working group document to come, based on draft-dnoveck-nfsv4-mv1-msns-update.   I've previously promised to submit that I-D by the end of August and expect to be able to do that.  
> These are listed as two separate milestones because the working group has not made any decision yet about how to go forward to address v4.1 issues with regard to trunking and state migration.  If the working group decides that these will be addressed in a single document, as I believe is best, these two milestones can be coalesced.  
> One possible occasion for the working group to make this decision is when a proposal is made to convert this document to working-group status.  Another is to address this as part of the milestone review.  In any case, there is no need to rush this decision since we could maintain these as two milestones indefinitely.
> Now let's look at the documents for the 2019 milestones:
> The first 1Q2019 milestone is for a WG document deriving from draft-cel-nfsv4-rpcrdma-cm-pvt-msg.  Although that document's current proposed status is standards-track, Chuck has decided that the milestone is for an anicipated informational document.  The working group can discuss the issue of the document's status as part of the milestone review or when Chuck formally asks to turn this into a working group document.  I understandthat there may be controversy about the  issue, but I don''t anticipte difficulty arrving at a consensus everyone can live with.
> The other 1Q2019 milestone is for a WG document deriving from draft-hellwig-nfsv4-rdma-layout.  As the discussion at IETF99 showed, there are still significant issues to resolve before this document will be complete.  I think we will want to identify those as part of the milestone review, but it is not necesssary to address them now, or even when the document becomes a WG document.  For those who think that Christoph's date is optimistic/aggressive, I agree but feel that given the guidance we have gotten from Spencer D., optimistic milestones are OK.  If, on the other hand, you feel that Christoph has a date which is unachievable/delusional, we can discuss that issue as part of milestone review.  If we do push that date off and Christoph does get this done by 3/31/2019, then he gets to tell all of us, "I told you so." 
> The 3Q2019 milestone, is for a WG document deriving from draft-cel-nfsv4-rpcrdma-version-two, of which I am a co-author.  There are still some technical issues to resolve, principally about the credit mechanism, but those do not need to resolved as part of the milestone review.  However, they do need to be identified so that we can go on to resolve them as the document is refined further over the next two years. 
> <NearFinalMilestoneDraft.pptx>

I have only very minor quibbles with any of this content, none worth
bringing up. This captures the bulk of what is ready to move forward,
and keeps the rate of publication roughly the same as its been for
the past five years. It's a good representation of the work before
us and the intent described in the proposed new Charter.

I suggested privately, and would like to bring up on the list, that
it would also be helpful to have a repository for our "stretch goals"
and "almost ready" or "wacky" ideas (thanks Bill Baker for that term).
It could help create an institutional memory, along with the mailing
list and IETF meeting notes (hint hint). We obviously have some worthy
ideas, as noted by Tom's "Next steps" presentation during IETF 99,
even if they aren't completely formed.

Speaking of institutional memory, I note that .pptx attachments do
not seem to be archived by either of the IETF's mailing list archive
tools. In one case, the attachment doesn't appear at all, and in the
other it is archived as a compressed file that doesn't seem to
re-inflate to anything useful.

Can you post subsequent revisions of the Charter and milestone list
only in ASCII text? That also makes it possible to simply hit "Reply"
and respond narrowly to individual parts of the text. Thanks!

Chuck Lever