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

Re: Can wavelength availability information be carried by OSPF-Ext for GMPLS?



Hi Shu,

Yes, in principle, whenever event of link-state change occurs, optical link information is updated.
How it causes a OSPF convergence problem depends on the frequency of the link-state update and 
network topology. 
I think that the wavelength availability information is needed to perform traffic engineering 
in optical networks. 
I am now describing this issue and the updated version will be posted soon.

Hope this helps. Best regards.
Eiji

On Sun, 9 Jun 2002 23:06:22 +0800
"#SHEN SHU#" <PG03053527@ntu.edu.sg> wrote:

> Hi, Eiji Oki,
> 
> Regarding your ID (draft-oki-ipo-optlink-req-00.txt), I have the following question:
> 
> How is flooding of optical link information originated? Is it triggered by event of link change? 
> --If yes, won't the frequent wavelength availability change impact OSPF convergence?
> --If no, how often will the timer be? Use the OSPF 30min max timer? how could up-to-date information garrentteed or approached?
> 
> I am not quite sure if such details should be included in your ID or stated in another document.
> If it's out of your scope, is there any other doc takeing care of this?
> 
> Your kindly answers/clarifications will be highly appreciated.
> 
> Best regards, 
> Shu SHEN 
> Research student 
> Network Technology Research Center 
> Nanyang Technological University, Singapore 
> Personal Website: http://shenshu.myip.org/
> 
> 
> -----Original Message-----
> From: #SHEN SHU# 
> Sent: Sunday, June 09, 2002 10:56 PM
> To: 'Ravi Ravindran'
> Cc: 'ccamp@ops.ietf.org'; 'oki.eiji@lab.ntt.co.jp'
> Subject: RE: Can wavelength availability information be carried by OSPF-Ext for GMPLS?
> 
> 
> Hi, Ravi,
> 
> I don't quite understand why you say information such as wavelength availability is "strictly binds to data plane".
> Is this because of wavelength availability is specific to LSC? or maybe a LSC from some vendors never use such information to make routing decisions?
> Either reason I would not agree.
> I think support for exchanging such information should be included within control plane, at least as an option.
> Please explain further.
> 
> About the ID you suggested, do you know what's its current status? I found it's not on the IPO WG's ID list.
> May be the authors of this ID can provide more information about it. 
> 
> Thanks a lot!
> 
> Cheers,
> Shu
> 
> -----Original Message-----
> From: Ravi Ravindran [mailto:rravindr@nortelnetworks.com]
> Sent: Sunday, June 09, 2002 4:27 AM
> To: #SHEN SHU#
> Subject: RE: Can wavelength availability information be carried by OSPF-Ext for GMPLS?
> 
> 
> OSPF-GMPLS extensions only try to address the common rourting issues relating to all non-packet switched interfaces. Extensions to things like you stated is more vendor specific, since that strictly binds to data plane property. 
> The following ID discusses some ways to deal with LSC nodes. 
> http://www.ietf.org/internet-drafts/draft-oki-ipo-optlink-req-00.txt 
> Ravi 
> -----Original Message----- 
> From: #SHEN SHU# [mailto:PG03053527@ntu.edu.sg] 
> Sent: Saturday, June 08, 2002 3:36 AM 
> To: ccamp@ops.ietf.org 
> Subject: Can wavelength availability information be carried by OSPF-Ext 
> for GMPLS? 
> 
> 
> Dear All, 
> Regarding draft-ietf-ccamp-ospf-gmpls-extensions-07.txt, my understanding is that there's no way to exchange wavelength availability information with such extension. Is this right?
> If a LSC router wants to make routing decision based on wavelength availability, how could it collect such information from each link of the network?
> Or does the OSPF extension leaves wavelength assignment to LDP completely? 
> Please comment, thanks! 
> Best regards, 
> Shu SHEN 
> Research student 
> Network Technology Research Center 
> Nanyang Technological University, Singapore 
> Personal Website: http://shenshu.myip.org/ 

--
Eiji Oki
NTT Network Innovation Laboratories
3-9-11 Midori-cho Musashino-shi, Tokyo 180-8585 Japan
TEL: +81-422-59-3441  FAX: +81-422-59-6387
E-mail: oki.eiji@lab.ntt.co.jp