The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2000-Oct> msg00212



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

(Reply) Hi seenu

  • From: seenu@samsung.co.kr
  • Date: Thu, 19 Oct 2000 00:04:01 +0900
  • X-OpenMail-Hops: 2

Hi mohan,

Sorry for a late reply.

Take the case  where an LSR has sent a No_Lbl_Resources notification to the peers that have requested for a label.
When Labels are available the LSR sends first a Lbl_Resources_Available notification and then sends a Label_Mapping msg.
See in this case, instead of sending two messages seperately, y can't u include the Status code ( Lbl_Resources_Available) in the
label mapping msg itself, which takes less time compared to the earlier one (since processing of a msg takes more time than a TLV).
Anyway if an LSR is not ready to handle that tlv it can always ignore it (draft-notifocation msg)

Take one more case : An upstream peer has detected a loop, so first it will send a loop_detected_noti and then sends a Lbl_release msg.
In this case also, u can include Status_tlv in the Lbl_Release msg

So depending on the situation it will reduces the computation.

Hope it will help u.

regards
Seenu

>       In the ldp-11 draft ,it's given that the status TLV
>  can be included in ldp messages which need not be
>  notif.
>      I feel there is a situation like the LSP got preempted,
>   which is a defined new status code, in such case it's helpful.
  
>      But do u think it's helpful in adding the status TLV 
> which ceratinly increases the processing at each node.


  • Follow-Ups: