|
Subject |
Draft minutes of 4th Tech WG meeting |
|
From |
Kevin Meynell <kevin@xxxxxxxxx> |
|
Date |
Tue, 29 Nov 2005 22:24:17 +0000 |
GLOBAL LAMBDA INTEGRATED FACILITY TECHNICAL WORKING GROUP
Minutes of the 4th meeting of the GLIF Technical Working Group held on
the 30th of September 2005 at Calit2, San Diego, United States.
Kevin Meynell - Issue 1
PRESENT
Name Organisation Country
---- ------------ -------
Tomonori Aoyama U. Tokyo Japan
Bill St. Arnaud CANARIE Canada
Erik-Jan Bos (Co-Chair) SURFnet The Netherlands
Greg Cole GLORIAD United States
John Connolly U. Kentucky United States
Steve Corbato Internet2 United States
Steve Cotter Internet2 United States
Sergi Figuerola i2CAT Spain
David Foster CERN -
Pat Gary NASA/GSFC United States
Vikram Gazula U. Kentucky United States
John Graham UKLight United Kingdom
Eduard Grasa i2CAT Spain
Leon Gommans U. van Amsterdam The Netherlands
Jan Gruntorad CESNET Czech Republic
Jeroen van der Ham U. van Amsterdam The Netherlands
René Hatem (Co-Chair) CANARIE Canada
Akira Hirano UIC/EVL United States
Hideaki Imaizumi U. Tokyo Japan
Masahiko Jinno NTT Japan
Akira Kato WIDE Project Japan
Scott Macdonald e-side Japan
Joe Mambretti iCAIR/Northwestern U United States
Steve Meacham NSF United States
Kevin Meynell (Sec) TERENA -
Dan Nae Caltech United States
Naohide Nagatsu NTT Japan
Bram Peeters SURFnet The Netherlands
Ana Preston Internet2 United States
Jan Radil CESNET Czech Republic
Predrag Radulovic U. Tennessee United States
Ann Richeson Qwest United States
Paul Roberts U. Houston TLCZ United States
Fay Sheu NCHC Taiwan
Matt Schmitz Cisco United States
Hideaki Tanaka KDDI R&D Labs Japan
Steven Thorpe MCNC United States
Christian Todorov Internet2 United States
Vasily Velikhov RRC Russia
Alan Verlo StarLight/TransLight United States
Steven Wallace Indiana University United States
Tom West National LambdaRail United States
Kennard White Glimmerglass United States
Garrut Yoshimi U. Hawaii United States
Oliver Yu UIC/EVL United States
[This meeting followed-on from the joint session of the Technical and
Control Plane Working Groups earlier in the day.]
1. ACTIONS FROM LAST MEETING
20050213-1 Kevin Meynell to change name of mailing list to
'tech@xxxxxxx'.
- Done.
20050213-2 Erik-Jan Bos, René Hatem, Cees de Laat, Jerry
Sobieski & Linda Winkler to write their sections for the
GLIF BCP by 15 March 2005.
- Ongoing.
20050213-3 Kevin Meynell to chase-up contributions and put
together draft document.
- Done, but insufficient contributions.
20050213-4 Jerry Sobieski to present final document at iGRID
2005.
- Superseded.
20050213-5 René Hatem to prepare some draft definitions for
discussion.
- Done.
20050213-6 Erik-Jan Bos to work on standardising service
descriptions.
- Done.
20050213-7 Linda Winkler and Jerry Sobieski to put
procedural information for SC'05 on the web before June
2005.
- Done.
20050213-8 Erik-Jan Bos to investigate the suitability of
the semantic web for cataloguing GLIF resources by 15
April 2005.
- Done.
20050213-9 René Hatem to clarify GLIF resources by 15 April
2005.
- Done.
20050213-10 Linda Winkler to send iGRID 2005 timelines and
milestones to Kevin Meynell by 6 March 2005.
- Done.
20050213-11 Kevin Meynell to put iGRID 2005 timelines and
milestones on the GLIF website.
- Done.
20050213-12 Kevin Meynell to establish links to iGRID 2005
and SC'05 from the GLIF website.
- Done.
20050213-13 Cees de Laat to start discussion on definition
of open and neutral optical exchanges on the mailing
list.
- Ongoing.
2. OPEN EXCHANGE COORDINATION
Erik-Jan said the communication between the optical exchange
operators (collectively known as GOLE) needed to be improved, and he
suggested to start holding regular teleconferences. These could be
organised by the GLIF Secretariat on (say) a monthly basis, although
a few logistical issues needed to be resolved.
There was general support for monthly teleconferences, although it
was felt that a chair should be appointed to put together an agenda
for each meeting. Christian commented that it was not reasonable to
burden one person with this task indefinitely, and thought the chair
should rotate on a quarterly basis, with each volunteer being
responsible for three meetings. He offered to take the first stint
as chair, which was accepted by everyone present.
The logistical issues were discussed, and it was agreed that each
meeting should generally not exceed one hour. Another problem was
finding convenient meeting times as participants were spread over
several time zones. It was agreed the times should be determined by
the chair, with the aim of minimising awkward dial-in times for as
many participants as possible. Erik-Jan mentioned that he had a
scheme for intercontinental meetings that he would send to the
mailing list.
ACTION 20050930-1: Erik-Jan Bos to send scheme for intercontinental
meetings to the mailing list.
It was agreed that the first teleconference should focus on
preparations for SC'05, so it needed to be arranged before then. The
GLIF Secretariat was asked to find a suitable date and set this up.
The participants would be HKLight, KR-Light, MANLAN, NetherLight,
NorthernLight, Pacific Northwest GigaPoP, StarLight, and T-LEX.
ACTION 20050930-2: Kevin Meynell to suggest suitable date for first
GOLE teleconference, and set-up this up before 24 October 2005.
The GLIF Secretariat was also asked to set-up a mailing list for the
optical exchange operators.
ACTION 20050930-3: Kevin Meynell to set-up GOLE mailing list.
3. FAULT MANAGEMENT
René said that it was becoming increasing important to implement an
end-to-end fault management system for GLIF. Although the majority
of the constituent networks were not production oriented, there
needed to be a centralised mechanism to report and track faults. For
that, a standard way of reporting faults needed to be agreed, and a
ticketing system implemented.
John agreed that a ticketing system should be developed, but he
thought existing practices should first be documented as part of the
BCP document. René therefore said he would circulate the CA*net4
Fault Management Guide as a starting point.
ACTION 20050930-4: René Hatem to circulate CA*net4 Fault Management
Guide on the mailing list.
4. RESOURCE SCHEDULING
Erik-Jan said this agenda item was related to the GLIF repository
discussed during the joint session of the Technical and Control
Plane Working Groups, although that was a longer-term activity. In
the meantime, it was important to collect as much contact, resource
and policy information from the GLIF operators as possible. This
would be made available on the GLIF website to help facilitate
coordination.
ACTION 20050930-5: Erik-Jan Bos to develop GLIF resource template.
ACTION 20050930-6: GLIF operators to send contact, resource and
policy information to the mailing list.
5. DEVELOPMENTS IN TRANSCONTINENTAL ETHERNET
John Graham gave a presentation on the use of Ethernet in the wide-
area (see
http://www.glif.is/meetings/2005/tech/graham-ethernet-over-
sonet.pdf). This considered Ethernet over MPLS tunnels and L2TPv3,
as well as over circuit-switched WANs (via ATM, POS and LAPS). In
particular, the new ANSI and ITU-T standards for GFP (Generic
Framing Procedure) with VCAT (Virtual Concatenation) and LCAS (Link
Capacity Adjustment) were examined.
GFP defines mapping for many types of service onto SONET/SDH or OTN,
including Ethernet, IP, Gigabit Ethernet, Fibre Channel and DVB
(Digital Video Broadcasting). It offers excellent bandwidth
utilisation, and efficency tailored to suit different client types.
It has simple delineation and robust error control, and as well as
being extensible.
The rationale behind GFP was that the cell overhead in traditional
ATM networks takes 10% of the bandwidth, and its adaptation
functions are overly complex. Similarly, Packet-over-SONET (POS)
requires all frames to be converted to PPP over HDLC, byte stuffing
causes non-deterministic bandwidth inflation, and QoS is hard to
monitor or guarantee. By constrast, GFP overs minimal overhead, it
can transport client PDUs (Protocol Data Units) in native format,
frames from multiple protocols can be easily aggregated into shared
bandwidth channels, and the low-latency capabilities are useful for
DVB and SAN (Storage Area Networking).
There are two modes of GFP: GFP-F which maps each client frame onto
a single GFP frame, and GFP-T which allows mapping of multiple
client frames onto 64B/65B code blocks for transport within a single
GFP frame. GFP-F offers higher bandwidth efficiency, but the
downside is higher latency, the need for more buffering, and the
need to calculate core header fields. By contrast, GFP-T supports
many protocols and offers low latency, but it is less bandwidth
efficient and requires more processing logic.
IEEE 802.3ae is the specification for 10 Gbps Ethernet over optical
fibre (i.e. there is no copper interface). This offers a 802.3 MAC
and frame format, although jumbo frames are not included in the
standard. It is full duplex only, does not support shared media, and
there is no CSMA/CD).
It is planned to implement these technologies on the UKLight's Ciena
CoreDirector and MetroDirector switches. This should simplify the
connections with other optical networks, and in particular remove
the need for a Cisco ONS 15454 between UKLight and the Nortel HDXc
at NetherLight.
Erik-Jan asked John whether his presentation could be turned into a
wider tutorial on Ethernet in the wide-area as people had found it
very informative. John replied that he would be willing to update it
and make it available on the web.
6. BEST CURRENT PRACTICES IN GLIF
Erik-Jan reported that progress on the GLIF BCP document had been
disappointing. A table of contents had been agreed at the previous
meeting, but contributions had only been received from John Graham
and René Hatem.
It was agreed that in order to progress this, perhaps just two or
three people with sufficient time should volunteer to write some
text. In anticipation of this, the table of contents should once
again be circulated to ensure that it was still relevant.
ACTION 20050930-7: Kevin Meynell to re-circulate BCP table-of-
contents.
7. NEXT MEETING
It had earlier been agreed that the next meetings of both the
Technical and Control Plane Working Groups would be held in
early-2006, in conjunction with an existing event.
[It was subsequently decided that the interim working group meetings
would be co-located with the Internet2 Joint Techs Workshop in
Albuquerque, USA; probably on 8-9 February 2006.]
OPEN ACTIONS
20050930-1 Erik-Jan Bos to send scheme for intercontinental
meetings to the mailing list.
20050930-2 Kevin Meynell to suggest suitable date for first
GOLE teleconference, and set-up this up before 24
October 2005.
20050930-3 Kevin Meynell to set-up GOLE mailing list.
20050930-4 René Hatem to circulate CA*net4 Fault Management Guide
on the mailing list.
20050930-5 Erik-Jan Bos to develop GLIF resource template.
20050930-6 GLIF operators to send contact, resource and policy
information to the mailing list.
20050930-7 Kevin Meynell to re-circulate BCP table-of-
contents.
