Re: Fw: [GSMP] more changes to the base draft

["Wang,Weiming" <[e-mail]>]

 Hi Avri,

Pls see inline. Thank you for reply.

Weiming

----- Original Message ----- 
From: "Avri Doria" <[e-mail]>
To: <[e-mail]>
Sent: Wednesday, November 26, 2003 7:54 PM
Subject: Re: Fw: [GSMP] more changes to the base draft


> Hi Weiming,
> 
> comments inline.
> 
> On onsdag, nov 26, 2003, at 17:52 Asia/Seoul, Wang,Weiming wrote:
> 
> >
> >  Hi Avri,
> >
> >  I'm writing to seek the possiblity in GSMP base header to add 
> > following items:
> >
> >  1. A proirity flag to indicate the processing priority for the 
> > message.
> 
> Can you explain how you would think this should be used?
> 

I suppose the timeliness for protocol messages like in ForCES may also be reqired by GSMP, and even, GSMP
may require this a little more, for GSMP needs to manage the connections from GSMP controller. A priority applying 
to some urgent messages like branch connection management may help in some cases.  

> >  2. An optional checksum part in GSMP for error control.
> 
> Again, can you explain how you would want to see this used?
> 
Also as being considered in GRMP, GSMP may sometimes be encapsulated on mediums that themselves cannot 
supply error control, like directly on Ethernet and ATM.  I also think GSMP is more probable to be encapsulated 
over such transport layer. Althougt GSMP may deal with the transmission errors in this case via the message field 
consistency, I just think a checksum may deal with this more specifically and easily. The idea to add the checksum is
the same as in UPD, only where a all zero of checksum is used to turn it off. When we set a flag for it, the bits can be saved
when checksum is turned off.

> >
> >  If possible, I also ask to leave some space in the Version field for 
> > possible including of other
> >  possible protocols like GRMP 
> > (http://www.ietf.org/internet-drafts/draft-wang-forces-grmp-01.txt).
> 
>   If the two protocols were to be merged it would seem to make the most 
> sense to me for there to be one base protocol which supported the MPLS 
> and GMPLS mode as well as the router mode.  I would personally be 
> uncomfortable with one version of the protocol for switches and one for 
> routers.  In many respects I think that the support  one would need for 
> adding messages for router support already exists.
> 
> If this merge is something that is of interest to the Wg and something 
> we can get IESG support for (i think it would require a charter 
> amendment) then I think we could certainly work out the details here on 
> the list.

Actually how a multiservice architecture is defined is important toward above
question. I do hope more progress can be made in MSF to help this. 

> 
> I do think your first 2 requests can be discussed within the current  
> work without any charter  problems.  For the other, I would like to 
> hear from other list members on whether it is something that the WG 
> should pursue.
> 
> a.
> _______________________________________________
> GSMP mailing list
> [e-mail]
> https://www1.ietf.org/mailman/listinfo/gsmp

#™¨¥Šx%ŠËFHÃâz×è®m¶›?ÿ0Ö'­~Šàþf¢–f§þX¬¶)ߣø,š