20231106-nfsv4

IETF-118 NFSv4 Agenda
Monday November 6, 2023
Welcome (back) to Prague
	•	Chairs Welcome / Agenda Bash (chairs)
	•	Working group progress (chairs)
	◦	document update
	◦	engagement
	•	Bakeathon report (??)
	•	RFC5661bis, Security, Internationalization (Noveck)
	•	Way forward

(1) Chairs Welcome / Agenda Bash (chairs)

Note well, Note Really Well (be nice) and agenda bash by Chris.

A helpful Meeting Tips slide for tool usage for remote participation.

(2) Working group progress (chairs)

(a) document update

Please see detailed document status slide from Chris.

delstid awaiting author update.

Layout WCC needs review on list to move through working group last call. There were no comments on the WG mail list. Restart WGLC. Short document, looking for a thumbs up.

rfc5661bis and internationalization docs to be discussed at this meeting.

rpc rdma-version-two update expected after November. 

(b) engagement (increase)

Chairs can use quiet moments between meetings to push people on document reviews and feedback. Talpey suggests regular messages to the working group lists.

Chairs need to circle back with the document authors to respond to AD review - Zahed.

(3) Bakeathon report

Chris summarized Bakeathon (NFS interoperability testing hosted by Redhat) from summary on mail list.

Talpey gave a live report on the Bakeathon. In mid-October about 12 attendees - half on site showed up at the RedHat facility in Westford, Massachusetts, with half testing via VPN. Primarily a test event, but discussions about protocol happened. 

Talpey was there to test RDMA capability. 

Dave Noveck mentioned that there are changes coming to the protocol and we may need follow up with Rick Macklem (FreeBSD) on the list - he did a presentation on directory delegation and questions about behaviour when recalling a delegation. The feeling was the spec wasn’t clear enough.

A Version 3 discussion ensued on handling timestamps in Linux - cache invalidation is happening too often - the timestamping is happening and unpredictable intervals (???). It was felt it was not a protocol issue but more an implementation issue.

Brian Pawlowski stated the Bakeathon NFS testing event is not an IETF event (just for clarity).

(4) RFC5661bis, Security, Internationalization (Noveck)

Dave Noveck presented -bis status. For the -02 version of doc there are five remaining errata reports. And reorganization of treatment of attributes.

<bis comments to be update>

<security comments to be updated>

<internationalization comments to be update>

(5) Way Forward

Request to start interim meetings every two weeks. (Chris scheduled these right away and the first is happening on Tuesday Nov 21).