********************************************************************** IETF 101 PALS - Monday, 19 March 2018 - 15:50-17:20 Room: Park Suite 50/90 min allocated; ** Please note the slot placement may be adjusted.) ********************************************************************** Chairs: Stewart Bryant and Andy Malis Secretary: David Sinicrope (x = slide sets NOT received as of 19 March 2018 17:40 (London time) 1. 15 min - Agenda bash, WG Agenda and Status - Andy MALIS and Stewart BRYANT Andy went through the slides. (It was noted that this could possibly be the last PALS WG meeting) Andy emphasized the new Note Well. There were no changes to the agenda. Andy called out the work in MPLS WG on draft-nslag-mpls-deprecate-md5 LDP still specifies use of MD5 for authentication. Sec Area is not in favor of MD5 and recommends strongly that it be updated. BGP has been using TCP-AO as a replacement for MD5. We are trying to have LDP follow what was done for BGP. Need to figure out cryptographic mechanism. Also it was noted there is an issue that there has not been a good deal of adoption of TCP-AO (as in not one implementation). Will coordinate with Sec Area and other Rtg Area WGs. It may be that the operators don't care and use MD5 as checksome and the Security ADs may not have understood this. MD5 is not good enough find another, like TCP-AO including all the specification that goes with it. Operator perspective seems to be that they use MAC Sec to secure all these things anyway, so MD5 is good enough. So long as drafts are not held hostage, would be a sufficient solution. Lou Berger: MAC Sec and single hop security for things going multiple hops doesn't work. Premise was that it was only used on one hop. There is some precedent for using something other than MD5 or TCP-AO. Yaakov Stein: Should not use MAC Sec as a substitute. These things happen at different layers. We may want a politically correct solution, but MAC sec would not be it and don't go with something inferior even if political. Andy: we will continue to work on an agreeable solution through the MPLS WG. 2. 20 min - Use of Ethernet Control Word RECOMMENDED - Stewart BRYANT https://datatracker.ietf.org/doc/draft-ietf-pals-ethernet-cw Objective: Calling out potential misordering issue with sending Ethernet packets in PWs with out the control word. Stewart went through the slides. Included are a number of changes made to the document since Singapore. (It was noted that the diagram in the WG slides, is for the presentation and is not included in the actual text.) Yaakov Stein: there is a better way to say some of the changes. Yaakov was encouraged to make any editorial changes to the PALS list. Yaakov: Can someone explain the objective Pat: While there is sympathy for the statement that was too strong about recommending the control word, the current text is far to week. Stewart: There were vendors that felt there was a sufficiently large deployment of the equipment that looked beyond the control word that didn't justify a stronger statement. Yaakov: this isn't saying anything by saying be careful. They should be careful anyway. If an implementation does not follow the IETF RFC then they do so at their own risk. Matthew: Saying anything stronger than be careful is something you would put in a user guide. Pat: if there was text saying that this shoudld only be enable if the network was designed to withstand it. Need to say what they need to be careful about. Yaakov: If this is informative, you could say that there are implementations that do such and such. You could say it is strongly recommended you don't. Matthew: Turn it around so that we are commenting on the behavior of the control word not what happens when you dont use it. Stewart: we were happy with the text up to the statement about misordering being hard to diagnose. Yaakov: its hard to diagnose an therefore highly recommended not to do it. Matthew: but you don't know the use case so you can't say its highly recommended not to do it. Andy: we could say "Operators should be aware .... this may cause Ethernet frames to be delivered out of order" Matthew: add ", depsite the presence of the PW CW." i.e., the CW is not a silver bullet for the first 4 bits being zero Yaakov: or that you could use the sequence number Stewart: that noone implements or uses Yaakov: we implemented the sequence number Andy: the proposal is that Stewart submit the text from the meeting as the last call text. Andy will start the last call and Mathew will judge consensus on the last call. There was no objection in the room to the proposal. The rest will be handled via the email list. 3. 15 min - Liaison about the changes in the control word draft - Andy MALIS N/A Objective: Notify those using Ethernet PWs in their work about the CW draft, and request it be taken into account in prior and future work. - Broadband Forum - ITU-T SG15 WP3 - IEEE 802.1, RAC - MEF Andy introduced the liaison. There was a question about how to address the RAC. Pat: send it to Angela (RAC Administrator) and ping Glenn Parsons or I for the address. Ieee-registration-authority@ieee.org (RAC Administrator) Andy: will send to Angela for redundancy Yaakov: and also RAC-Chair@IEEE.org Andy went through the text of the liaison taking comments in real-time. Pat: trying to make the addresses last, not running out, but not going to throw out addresses Andy: damage is already done and nothing the RAC can do so we need to fix it. There were no wording changes proposed except to delete "Committee" from the IEEE Registration Authority and to fix the acronym. Deborah: need to be more definitive about when to send in comments. Not sure why they would need to comment, they would just need the information. Yaakov: There was some speculation that SG15 made the CW mandatory. They did not Dave: There are other forums that have it as optional. Andy: I'll send the liaison when v04 is online from Stewart Andy: Any other business? Hearing none, we are adjourned. ********************************************************************** Overflow (Will be presented if time permits.) ********************************************************************** xx. - None currently ********************************************************************** REMOTE INFORMATION FOR THE PALS SESSION(S) ********************************************************************** Remote Participation Info: http://www.ietf.org/meeting/101/remote-participation.html (link not yet active) - No WebEx - IETF 101 Agenda with Audio and Jabber links: https://tools.ietf.org/agenda/101/ (link not yet active)