[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: GMPLS Last Calls
Dear Jerry and Percy,
I understand you concerns very well. Of course the restoration is of
paramount importance and all GMPLS co-authors are convinced about that. We
didn't forgot it, indeed we specifically decided to not include it at this
stage.
The process is first (if possible) to define an MPLS based restoration and
then to generalize it for GMPLS. This is still this idea of re-engineering
that we are doing since the beginning.
Another point is that this is very touchy and requires a lot of additional
thinking. Also if you consider the amount of religious debate that we
already have for SDH/SONET transparency and concatenation, I let you guess
what it will be for restoration (the IETF defining protection/restoration
protocols applicable to SDH/SONET.... we will have a lot of fun on mailing
lists :-)
The GMPLS specification in WG last call today is the first version. We
didn't want to include and cover everything. Additional work is still
needed. In some cases it should first fit the MPLS concepts and then go to
GMPLS.
Kind regards,
Eric
-----Original Message-----
From: Tarapore, Percy S [mailto:tarapore@att.com]
Sent: Thursday, May 24, 2001 8:55 PM
To: Ash, Gerald R (Jerry); ccamp@ops.ietf.org; mpls@uu.net
Subject: RE: GMPLS Last Calls
There is a significant issue related to the absence of restoration in the
signaling draft
(http://www.ietf.org/internet-drafts/draft-ietf-mpls-generalized-signaling-0
4.txt). Very specifically, Section 7 of this draft deals with the case
involving protection information for link protection. While link protection
schemes may be desirable for fast recovery related to high priority LSP's,
more cost-effective shared mesh restoration schemes would be preferred for
the majority of traffic from a Service Provider's perspective. This
observation is supported by the fact that many vendors are currently
developing proprietary schemes for shared mesh restoration. Hence, in
addition to the protection information, GMPLS signaling needs to reflect a
minimum set of information/attributes required for shared mesh restoration
without jeopardizing vendor proprietary solutions.
The need for multiple types of restoration capabilities is well documented
in the OIF/UNI I-D
http://search.ietf.org/internet-drafts/draft-many-carrier-framework-uni-00.t
xt as follows:
" Multiple types of facilities available for restoration are needed
within the network. The following options should be considered for
allocation of facilities to support restoration of failed
connections:
- Dedicated restoration capacity
- Shared restoration capacity. This allows the network to ensure
high quality service for customers, while still managing its
physical resources efficiently.
- Un-restorable
- Pre-emptable"
The OIF/UNI I-D supports a range of different restoration schemes through
the use of service level as a connection attribute. This attribute is
defined as follows:
" an integer attribute that indicates a class of service. A carrier may
specify a range of different classes of service (e.g. gold, silver, bronze)
with predefined characteristics (e.g. restoration plans). The pre-defined
service types correspond to different types of network restoration (e.g. no
restoration, 1+1 protection), connection set-up and hold priorities,
reversion strategies for the connection after failures have been repaired,
and retention strategies."
It is therefore important that GMPLS be extended to be able to support such
restoration schemes.
Percy S. Tarapore
AT&T Labs
-----Original Message-----
From: Ash, Gerald R (Jerry)
Sent: Thursday, May 24, 2001 2:20 PM
To: ccamp@ops.ietf.org; mpls@uu.net
Cc: Ash, Gerald R (Jerry)
Subject: RE: GMPLS Last Calls
Quoting from the CCAMP/IETF-50 meeting minutes re the GMPLS Architecture
draft:
"Eve - Had hoped that CCAMP formation and expression of requirements would
enable us to do more with architecture than reverse architect proposed
solution. Think we should also look at carrier requirements and look for
discrepancies with architecture.
Curtis - requirements of carriers not being addressed?
Eve - happy to discuss on mailing list in absence of time"
There are still no documented service provider (SP) requirements driving the
proposed GMPLS protocol extensions. This is inconsistent with the current
initiatives to provide SP requirements prior to protocol extensions being
accepted, such as for protection/restoration, network hierarchy, MPLS OAM,
MPLS/DiffServ TE, multi-area TE, etc.
Here is a sample of SP requirements that are not being addressed (other
requirements from our perspective are forthcoming):
1. Restoration requirements, particularly in support of mesh restoration,
need to be supported by GMPLS. For example, Section 7 of the Generalized
Signaling I-D
(http://www.ietf.org/internet-drafts/draft-ietf-mpls-generalized-signaling-0
4.txt)
primarily discusses link protection, restoration capabilities are largely
missing and need to be included.
2. Standards explicitly supported by GMPLS, such as G.707, G.709, etc.,
should be clearly identified in the text and referenced in the GMPLS I-Ds,
e.g., in Section 3.1 of the Generalized Signaling I-D, add "G.707 [Reference
G.707] is supported by the Generalized Label Request."
As per Eve's comment at IETF-50, SPs are encouraged to post their
requirements to the list. It would also help if more SPs were invited to
co-author the GMPLS drafts, to help ensure that SP requirements are more
adequately reflected and addressed.
Jerry Ash
AT&T Labs
-----Original Message-----
From: George Swallow [mailto:swallow@cisco.com]
Sent: Monday, May 14, 2001 1:52 PM
To: ccamp@ops.ietf.org; mpls@uu.net
Subject: GMPLS Last Calls
This message initiates a last call on four GMPLS drafts. The last
call is being issued jointly in the MPLS and CCAMP workgroups.
The drafts are:
1. Generalized MPLS - Signaling Functional Description
<draft-ietf-mpls-generalized-signaling-04.txt>
2. Generalized MPLS Signaling - RSVP-TE Extensions
<draft-ietf-mpls-generalized-rsvp-te-03.txt>
3. Generalized MPLS Signaling - CR-LDP Extensions
<draft-ietf-mpls-generalized-cr-ldp-03.txt>
4. GMPLS Extensions for SONET and SDH Control
<draft-ietf-ccamp-gmpls-sonet-sdh-00.txt>
The last call closes May 29, 12 pm GMT.
- V2KG (Vijay, Vijay, Kireeti, & George)
======================================================================
George Swallow Cisco Systems (978) 244-8143
250 Apollo Drive
Chelmsford, Ma 01824