[email protected]
[Top] [All Lists]

RE: composite link - candidate for respin, maybe

Subject: RE: composite link - candidate for respin, maybe
From: John E Drake
Date: Thu, 1 Apr 2010 07:45:00 -0700
> -----Original Message-----
> From: [email protected] [mailto:[email protected]]
> Sent: Wednesday, March 31, 2010 11:56 PM
> To: John E Drake
> Cc: [email protected]; Tony Li; Mcdysan, David E; [email protected]
> Subject: Re: composite link - candidate for respin, maybe
> 
> 
> In message <[email protected]>
> John E Drake writes:
> >
> >
> > > John,
> > >
> > > When nothing is being moved nothing is reordered.  If a flow or many
> > > flows is moved from a link with longer delay to one with shorter
> > > delay, unless the flow is stopped and bufferered for at least the
> > > delay difference and then drained, then reordering will occur.
> > >
> > > So I agree with Tony.  It is not practical to avoid reordering when
> > > traffic is moved.  It is therefore minimally disruptive.  The products
> > > of your new employer do the same thing.
> > [JD]
> >
> > One last time: Flows may be re-ordered, but packets within flows are
> > not re-ordered.  This is no different than placing flows of a single
> > LSP on multiple component links, which is a required behavior.  The
> > only difference is that when an LSP is pinned to a specific component
> > link, this only occurs when the LSP is being pinned to a new component
> > link using MBB.
> 
> 
> 
> John,
> 
> You are making no sense at all.
> 
> Consider the case where only one flow exists.  Two path exist, A and
> B.  Path A is used.  It has a longer delay than B.  The difference in
> delay is N packet times.  When switching from A to B, the first packet
> to arrive on path B arrives N-1 packets before the last packet to go
> into link B.  Packets from that one flow are interleaved for about N
> packets.
> 
> The case that we are talking about is moving a flow or group of flows
> from one link to another.  It is minimally disruptive.  Do you have a
> problem with that wording?
[JD] 

Fine, in the limiting case of an LSP with a very small number of flows, there 
could be packet re-ordering. 
 
> 
> Curtis

_______________________________________________
rtgwg mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/rtgwg

<Prev in Thread] Current Thread [Next in Thread>