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

Re: LCAS and GMPLS



Hi Maarten, thanks. Adrian and ITU folks, I went to the additional CCAMP web page and looked at the liasons but didn't see
G.806, G.783, G.783, G.798 nor G.7042 (LCAS itself) listed. Are these available to the CCAMP workgroup. My archive of ITU-T
documents is a bit out of date.


Stephen, did Maarten's answer clear things up? Think about VCAT as a physical layer (circuit based) inverse multiplexing technique, hence to
keep the hardware reasonable the spec. (G.707) doesn't allow for mix and matching component types in a VCAT group. Or did you mean something
else with your recursive composition idea?


Adrian, I've started on the I-D using the relatively new XML method (seems easier than the MS Word or pure text method). Anyone who wants to help
or review prior to submission let me know. Note that my gregb@grotto-networking.com e-mail address is checked more frequently than my gregbern@yahoo.com address.


Greg B.
Maarten.Vissers@alcatel.de wrote:

Greg,



E.g., what about renumbering?


The renumbering is a LCAS element (fully LCAS controlled), that NMS or GMPLS should not be bothered about.
Only if you want to operate VCAT with LCAS disabled, NMS/ASON/GMPLS has to control the sequence numbers, but changes in the group size is not hitless without LCAS.


Hereafter some information on the VCAT/LCAS associated parameters as extracted from G.806 (02/04) and G.783 (02/04).

10.1/G.806 includes a generic specification for the VCAT/LCAS functions, inlcuding a list of MI signals (MI: Management Information) which management needs to control and gets reported. G.783 and G.798 specify the technology specific functions, using the specifications in G.806 as basis.

Input MI signals:
=================
P-Xv/P-X-L_A_So_MI_LCASEnable:
The MI_LCASEnable input controls whether the LCAS functionality is enabled for the source function (MI_LCASEnable = true) or disabled (MI_LCASEnable = false).


P-Xv/P-X-L_A_So_MI_ProvM[1..XMT]:
The MI_ProvM[1..XMT] input controls whether a P[i]_AP at the P-Xv_AP is provisioned to be a member of the VCG (MI_ProvM[i] = 1) or not (MI_ProvM[i] = 0).


==> With this input parameter NMS/GNPLS controls which access points (and thus VC-n trail termination functions) are part of the VCAT group. Note - LCAS determines if there will be traffic routed over it.

P-Xv/P-X-L_A_So_MI_PLCTThr:
Partial loss of capacity transmit threshold associated with cPLCT fault cause.


----------

P-Xv/P-X-L_A_Sk_MI_LCASEnable:
The MI_LCASEnable input controls whether the LCAS functionality is enabled for the sink function (MI_LCASEnable = true) or disabled (MI_LCASEnable = false). If LCAS is enabled, the function auto-detects which type of source it is interfacing to. The output MI_LCAS_So_Detected reports whether the present sink function detected an LCAS-enabled source function (MI_LCAS_So_Detected=true) or a non-LCAS-enabled source (MI_LCAS_So_Detected=false, see below for details). Only if both MI_LCASEnable and MI_LCAS_So_Detected are true is the LCAS functionality active in the function.


P-Xv/P-X-L_A_Sk_MI_ProvM[1..XMR]:
The MI_ProvM[1..XMR] input controls whether a particular one of the available physical resources at the P-Xv_AP is provisioned to be a member of the VCG (MI_ProvM[i] = 1) or not (MI_ProvM[i] = 0).


==> With this input parameter NMS/GMPLS controls which access points (and thus VC-n trail termination functions) are part of the VCAT group. Note - LCAS determines if there will be traffic routed over it.

P-Xv/P-X-L_A_Sk_MI_PLCRThr:
Partial loss of capacity receive threshold associated with cPLCR fault cause.


P-Xv/P-X-L_A_Sk_MI_TSDEnable:
The MI_TSDEnable input controls whether the sink function uses AI_TSD[i] indications as contributors for signalling defective members back to the LCAS source function (MI_TSDEnable = true) or whether it ignores AI_TSD[i] indications altogether (MI_TSDEnable = false).


P-Xv/P-X-L_A_Sk_MI_HOTime:
The MI_HOTime input controls whether the Hold-Off (HO) timer is enabled or disabled for the sink function and, if enabled, what the value of the timer is. If MI_HOTime = 0, the HO timer shall be disabled, if MI_HOTime 0 it shall be enabled.


P-Xv/P-X-L_A_Sk_MI_WTRTime:
The MI_WTRTime input controls whether the Wait-To-Restore (WTR) timer is enabled or disabled for the sink function and, if enabled, what the value of the timer is. If MI_WTRTime = 0, the WTR timer shall be disabled, if MI_WTRTime 0 it shall be enabled. The range of values for the HO/WTR timers is as defined in ITU-T Rec. G.808.1.


P-X-L_TT_Sk_MI_SSF_Reported:
Controls if cSSF fault cause is reported. Typically owned by NMS.


Output MI signals:
==================
P-Xv/P-X-L_A_So_MI_XAT:
Current size of the transmitted payload.


P-Xv/P-X-L_A_So_MI_XMT
Maximum size of the transmitted payload.

P-Xv/P-X-L_A_So_MI_TxSQ[1..XMT]
Indication of which sequence number (_SQmap[i]) is being carried over a particular P_AI (P_AI[i]) signal.


P-Xv/P-X-L_A_Sk_MI_XMR
Maximum size of the received payload.

P-Xv/P-X-L_A_Sk_MI_XAR
Current size of the received payload

P-Xv/P-X-L_A_Sk_MI_DMFI[1..XMR]
The relative delay, in MFI units, between each provisioned member and the earliest-arriving member among those considered for the calculation


P-Xv/P-X-L_A_Sk_MI_LCAS_So_Detected
Indication if remote node supports LCAS.

P-Xv/P-X-L_A_Sk_MI_AcSQ[1..XMR]
Indication of which sequence number is being received over a particular P_AI (P_AI[i]) signal.



Besides these VCAT/LCAS specific parameters, you need to control a subset of the individual trail termination (e.g. S4_TT) function and the client adaptation (e.g. S4-X/ETH_A) function parameters. E.g. for VC-4 and VC-3 TT_Sk functions:


Input MI parameters:
====================
Sn_TT_Sk_MI_TPmode
Termination Point mode to control alarm free path set up. Refer to 6.1/G.806 and 7.1.4/G.7710.


Sn_TT_Sk_MI_ExTI
Expected Trace Identifier value must be configured by GMPLS when GMPLS sets up VC-n network connection; action is to read MI_TxTI parameter in TT_So and carry this value to remote end to configure MI_ExTI.


Sn_TT_Sk_MI_TIMdis
Disables/Enables TIM defect detection. GMPLS should control this as part of VC-n connection setup.


Sn_TT_Sk_MI_TIMAISdis
Disables/enables AIS insertion on TIM defect detection. GMPLS to control this as part of VC-n connection setup.


Sn_TT_Sk_MI_RDI_Reported
Sn_TT_Sk_MI_SSF_Reported
Controls if cRDI and cSSF fault cause are reported. Typically owned by NMS.


Sn_TT_Sk_MI_DEGTHR
Sn_TT_Sk_MI_DEGM
Sn_TT_Sk_MI_EXC_X
Sn_TT_Sk_MI_DEG_X
Bit error defect detection process configuration parameters. Must be configured as part of connection setup; values depend on SLA.



Output MI parameters: ===================== Sn_TT_Sk_MI_AcTI Accepted Trace Identifier value report.

Parameters for one of the client adaptation functions (e.g. Sn/ETH_A, G.8021) are typically outside the control of the GMPLS VC-n Group connection setup; those parameters are under either NMS or client layer GMPLS control.


I hope the above helps to get a feeling of what you need to control during VC-n Group connection set up.
Note that the VCAT group can be set up as a single VC-n Group, or as multiple, diverse routed, (smaller) VC-n (sub)Groups. E.g. a 5 VC-4 VCAT group may be set up as a 5 VC-4 network connection group, as a 3 VC-4 netowrk connection group (VCG A) and a 2 VC-4 network connection group (VCG B), as five individual VC-4 network connections, ... The most common case when used to transport data will most likely be the second case: two VCGs supporting one VCAT/LCAS group. I assume that NMS and GMPLS VC-n connection management would set up VCG A and a VCG B, with the requirement that both must be diverse routed.


If the VCAT group at a later stage is to be extended to 6 VC-4 members, then for the case of VCG A and VCG B, the VCG B could be extended from 2 to 3 VC-4 network connections. You would then issue a connection modification request for VCG B and leave VCG A untouched.

Regards,
Maarten









Greg Bernstein <gregbern@yahoo.com> Sent by: owner-ccamp@ops.ietf.org 15-06-2005 18:04

To: Adrian Farrel <adrian@olddog.co.uk>, Diego Caviglia <Diego.Caviglia@marconi.com>
cc: yhwkim <yhwkim@etri.re.kr>, ccamp <ccamp@ops.ietf.org>
Subject: Re: LCAS and GMPLS



Okay Adrian I'll take a first stab at the I-D based on
the e-mails to date and some of my tutorial material
from short courses and papers.


Thanks everyone for the info and particularly Maarten
for the in depth descriptions and information. I may
send you some extra questions after I digest it. I'm
particularly interested in places where we "enable
LCAS" or some entity tells LCAS to perform an "add" or
"delete" action. E.g., what about renumbering?


Sounds like there are plenty of possible solutions, so
a focus on a clear understanding of the problem space
and capabilities (which are already automated, can be
automated, or need to be commanded) is important.

Greg B.

--- Adrian Farrel <adrian@olddog.co.uk> wrote:
--- Lots of snips
I would suggest that Greg and Diego start an I-D. Call
it something like "Applicability Statement for
Operating LCAS and VCAT with GMPLS LSPs".
Include:
- Simple overview of VCAT and LCAS (no more than a
page, please)
- Simple statement of how LSPs fit into the picture
(about half a page)
- Statement of the requirements on GMPLS signaling
(about a page)
- Mechanisms and procedures (two or three pages)

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com