[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: SONET/SDH label agreement for IETF, ITU-T and OIF




Guys... I have seen to much of this. I have asked Kireeti
EXPLICITLY to try and CALL FOR or DECLARE CONSENSUS on the
WG mailing list. I do NOT want another 500 emails going back
and forth on this issue. We need to approach this pragmatically.

- WG Chair(s) try to get (rough) CONSENSUS CALLED OUT on the 
  WG mailing list on what exactly we agreed in SLC. That will
  help to prepare a response to ITU-T as well
- Based on that WG chairs try to get (rough) CONSENSUS CALLED OUT
  on the WG mailing list on the exact text. 
- Then Eric (editor) makes the changes as per CONSENSUS
- Then we're done.

Sorry that I need to step in, but this has been going on TOO long.

Bert 
speaking as AD

> -----Original Message-----
> From: Mannie, Eric [mailto:Eric.Mannie@ebone.com]
> Sent: Friday, February 22, 2002 5:03 PM
> To: 'ccamp@ops.ietf.org'
> Cc: 'sob@harvard.edu'; 'mvissers@lucent.com'; 'kireeti@juniper.net';
> 'vijay@umbc.edu'; 'bwijnen@lucent.com'
> Subject: SONET/SDH label agreement for IETF, ITU-T and OIF
> Importance: High
> 
> 
> Hi All,
> 
> Please note:
> 
> The two updated drafts "implementing" the agreement were sent 
> on the CCAMP
> mailing list the 14th of December 2001.
> 
> > move "Appendix 1 - Signal Type Values Extension For Group 
> > Signals" from the sonet-sdh document to the 
> sonet-sdh-extensions document;
> 
> Was done and sent on the CCAMP mailing list as said before. I 
> don't seen
> where the agreement was not fulfilled (just copy/pasted the 
> appendix to the
> non standard draft).
> 
> > Afterwards I noticed that the latter agreement is 
> > interpreted in different ways
> 
> The sentence that was added in the intro to implement the 
> agreement was
> reviewed by Maarten. It could certainly be improved, I agree, 
> but this was
> the result of several hours of discussions.
> 
> "A SONET signal which has an identical SDH signal SHOULD be 
> requested using
> the same traffic parameters as for the equivalent SDH signal, and will
> consequently use the SDH label."
> 
> The key point here is that the LSP Encoding Type (i.e. the 
> circuit type) in
> the root GMPLS signaling will indicate either SDH or SONET 
> (xor). It is
> impossible to have a double coding since we have an exclusive 
> OR. It is
> impossible to interpret it in two different ways.
> 
> The sentence above should better speak explicitly in terms of 
> LSP Encoding
> Type.
> 
> Having said that, there is something magic of having the 
> GMPLS SDH/SONET
> specification accepted also by the ITU-T as part of GMPLS for G.ASON
> (on-going activity). It would be a pity if the SONET label 
> format prevents
> it.
> 
> From my humble point of view, and since I understood that 
> ITU-T and ATM
> Forum have decided to work on PNNI extensions, there is also 
> something magic
> of using the same traffic parameters for GMPLS and PNNI, and 
> moreover at the
> IETF, ITU-T, ATM Forum and OIF. That will allow further 
> interoperability
> between the GMPLS and the PNNI control planes.
> 
> It would be a mistake if these discussions on the label 
> format prevent us to
> use the traffic parameter, and moreover if it results in 
> different label
> formats between IETF and ITU-T.
> 
> If we change the SONET label format there are two issues:
> 
> - already used by the OIF for UNI1.0.
> - already implemented in many boxes.
> 
> The OIF started to work on UNI 2.0 and we could suggest to 
> change the SONET
> label such as it is fully identical to the SDH label as part 
> of UNI 2.0.
> 
> Is there somebody fundamentally against changing the SONET 
> label format in
> the current context ?
> 
> I know that these issues are larger than the scope of IETF 
> but it becomes
> difficult to work without encompassing the broader scope of 
> also ITU-T, OIF
> and now ATM Forum.
> 
> I would suggest that Maarten and myself (with the help of Juergen and
> Dimitri Papadimitriou) review the two SDH/SONET drafts and 
> reach a FINAL
> consensus (as far as we are concerned) such as these 
> documents could be
> re-used in different contexts to ensure interoperability. I 
> am confident
> that by the end of next week we could republish these two 
> drafts with a very
> strong consensus.
> 
> Please advice.
> 
> Kind regards,
> 
> Eric
> 
> -----Original Message-----
> From: Wijnen, Bert (Bert) [mailto:bwijnen@lucent.com]
> Sent: Friday, February 22, 2002 3:13 PM
> To: Kireeti Kompella; Vijay Gill
> Cc: ccamp; Maarten Vissers; Scott Bradner
> Subject: RE: SONET/SDH label agreement?
> 
> 
> WG chairs, I think Maarten is correct here in the sense
> that by now, he (and the WG mailing list) could have 
> expected an answer. 
> 
> PLEASE act asap.
> 
> Bert 
> 
> > -----Original Message-----
> > From: Maarten Vissers [mailto:mvissers@lucent.com]
> > Sent: Friday, February 22, 2002 9:17 AM
> > To: Kireeti Kompella; Vijay Gill
> > Cc: ccamp; Wijnen, Bert; Scott Bradner
> > Subject: Re: SONET/SDH label agreement?
> > 
> > 
> > Kireeti, Vijay,
> > 
> > On Feb. 8 I send you the email below. So far I haven't 
> > received an answer and I
> > assume the email has not been received by you or has got at 
> > the bottom of the
> > stack. As such a resent. Hope you will be able to respond today.
> > 
> > Thanks,
> > 
> > Maarten
> > 
> > Maarten Vissers wrote:
> > > 
> > > Vijay, Kireeti,
> > > 
> > > Almost two months ago we met in a small team to address the 
> > issues hindering the
> > > completion of draft-ietf-ccamp-gmpls-sonet-sdh and
> > > draft-ietf-ccamp-gmpls-sonet-sdh-extensions.
> > > When this meeting ended, I was convinced we had reached 
> > agreement on the way to
> > > continue:
> > > 
> > > - move "Appendix 1 - Signal Type Values Extension For Group 
> > Signals" from the
> > > sonet-sdh document to the sonet-sdh-extensions document;
> > > 
> > > - modify the sonet-sdh document such that the SDH traffic 
> > parameters and label
> > > will be used for SONET signals for which there exists an 
> > identical SDH signal.
> > > SONET signals for which there is no SDH equivalent will 
> > keep using the SONET
> > > specific traffic parameters and label.
> > > 
> > > Afterwards I noticed that the latter agreement is 
> > interpreted in different ways:
> > > 
> > > A) keep both SONET and SDH specific traffic parameter and 
> > label specifications
> > > in the sonet-sdh document, and let the equipment 
> > manufacturer and/or operator
> > > choose if the traffic parameters and label for a SONET 
> > signal (with identical
> > > SDH signal) will use the SONET specification or the SDH 
> > specification. This
> > > results in a "double coding" scheme for SONET signals.
> > > 
> > > B) modify the sonet-sdh document such that there is one set 
> > of traffic
> > > parameters and label for each SONET signal. For those SONET 
> > signals with
> > > identical SDH signal (i.e. all SONET signals except VT-3) 
> > only the SDH traffic
> > > parameters and label will be specified. For those SONET 
> > signals that do not have
> > > an SDH equivalent (i.e. VT-3) the SONET traffic parameters 
> > and label will be
> > > specified. This results in a "single coding" scheme for 
> > SONET signals.
> > > 
> > > This dual interpretation is again hindering the completion 
> > of the sonet-sdh
> > > document.
> > > 
> > > Note that interpretation B) sufficiently meets the request 
> > from ITU-T SG15 as
> > > laid down in the its communications statement and as such 
> > was an acceptable
> > > compromise for me.
> > > 
> > ftp://sg15opticalt:otxchange@ftp.itu.int/tsg15opticaltransport
> /COMMUNICATIONS/ccamp/IETF_ccamp_sdhgroup.html
> > (note -  I can't find this document on the IETF web site anymore)
> > Interpretation A) will at the best require two coding schemes to be
> supported in
> > each equipment, and at the worst will cause interworking 
> problems. It
> doesn't
> > meet the request from ITU-T SG15. If I would have been aware of this
> > interpretation, I would not have agreed with it.
> > 
> > May I ask you for your understanding/interpretation on this matter.
> > 
> > Regards,
> > 
> > Maarten
>