Re: [mpls-tp] poll on making draft-sprecher-mpls-tp-survive-fwk-01.txtaworking group document?

["Xueqin WEI (Shuetsing WEI)" <[e-mail]>]

Huub:

Sorry updated, so I send again.
Thanks for you reply, and please see follow:

>Make-b4-break requires merging which is not supported.
I agree with you! In packet network, protection need merging at the destination end.

>Also make-b4-break may cause PDU duplication which
>cannot be detected.
I agree with you and I think duplicaiton problem is more important! Make-b4-break
 will cause bridge at the source end, and this will cause duplicaton. So we cann't 
support make-b4-break in a packet network (MPLS-TP) .

>The one (1) protection entity must be possiible to protect 100% one
>of the N working entities. This means that if the N working entities
>have different bandwidth, the protection entity MUST have the
>capability to protect the largest bandwidth in the N working entities
1:N protection means N work path share the bandwidth of 1 protection path. 
If 1 of N working path failed, then this working path will be 100% protected.
But when more than 1 working path failed, then only 1 of N working path will be 
protected and the rest failed working path will still failed. IMO, the second one 
can't be called 100% protection.


Sincerely Yours
Xueqin Wei (Shuetsing Wei)

Development & Planning Department,
Fiberhome Telecommunication Technologies Co.,Ltd.,
No.88 Youkeyuan Road,Hongshan Dist.,Wuhan,Hubei,P.R.China,430074
Tel:    [phone]
Fax:    [phone]
Email:  [e-mail]
2009-03-13  08:21:55

==================== Following is your email=====================
Subject:Re: [mpls-tp] poll on making	draft-sprecher-mpls-tp-survive-fwk-01.txtaworking group document?
Sent：2009-03-13 08:12:58
From:Huub van Helvoort
To:xqwei
CC to:MPLS-TP

>Ni hao Xueqin,
>
>You wrote:
>
>> Follows are several minor comments:
>> 
>> Make-before-break and break-before-make have no direct
>> relation with soft re-routing and hard re-routing. Of course, soft
>> re-routing should apply 
>> Make-before-break to reduce the effect to triffic.
>
>Make-b4-break requires merging which is not supported.
>
>Also make-b4-break may cause PDU duplication which
>cannot be detected.
>
>> 2, In section 3 for protection:
>> 
>>> - Must support protection ration of 100%.
>>
>> This is impossible for 1:N protection when more than 1 working path
>> need protect.
>
>The one (1) protection entity must be possiible to protect 100% one
>of the N working entities. This means that if the N working entities
>have different bandwidth, the protection entity MUST have the
>capability to protect the largest bandwidth in the N working entities
>
>Cheers, Huub.
>
>-- 
>================================================================
>Always remember that you are unique...just like everyone else...
>
>

=================================================================