Re: [BEHAVE] Comments on draft-chen-behave-rsnat-01

[Chen Gang <[e-mail]>]

Hello Dave,

The draft is to resolve similar problems in "Redundancy and Load Balancing
for Stateful NAT".

We look forward to your comments.

Thanks

BRs

Gang



2009/7/23 Dave Thaler <[e-mail]>

> This draft seems to be tackling the same problem as
> draft-xu-behave-stateful-nat-standby-00 except using
> slightly different protocols.
>
> Terminology
>
> Chen et al.       Xu
> ------------      --------
> User Network      internal realm
> Service Network   external realm
>
> Protocols
>
> Function                Chen et al.      Xu
> ---------------------   --------------   -------------------
> Liveness detection      BGP KEEPALIVE    VRRP over VPLS
> State synchronization   new BGP option   SCSP
> Routing preferences     BGP priority     More/less specifics
>
>
> I also have a bunch of editorial-only comments which I will
> send directly to the authors.
>
> -Dave
>
> > -----Original Message-----
> > From: [e-mail] [mailto:[e-mail]] On
> > Behalf Of Dave Thaler
> > Sent: Tuesday, July 21, 2009 7:28 PM
> > To: [e-mail]
> > Cc: Behave WG
> > Subject: [BEHAVE] Comments on draft-xu-behave-stateful-nat-standby-00
> >
> > This document is now much more clear than
> > draft-xu-behave-lsn-standby-00 was, thanks!
> >
> > I mostly just have some more editorial nits, which I
> > will send directly to you.  I only have one technical comment...
> >
> > Section 4.2 states:
> >
> > > To preserve the established sessions during the failover, in
> > > addition to keeping the internal addresses for the external hosts
> > > unchanged, the external addresses for the internal hosts should also
> > > keep unchanged. How to meet the first requirement will not be
> > > reiterated since it is the same as that for the cold standby
> > > mechanism. To meet the second requirement, NAT routers in a
> > > redundancy group should be configured with an identical external
> > > address pool and they should assign the same external address for
> > > the same internal host.
> >
> > It's not just the same address that a stateful NAT has to assign,
> > it also has to assign the same external port.  So the state
> > synchronization protocol also has to exchange port mapping
> > information.  In Section 7 you suggest using SCSP (RFC 2334), which
> > would require you to specify the record format (see section B.2.2)
> > which would need to include both addresses and ports, and perhaps
> > other information as well.
> >
> > -Dave
> > _______________________________________________
> > Behave mailing list
> > [e-mail]
> > https://www.ietf.org/mailman/listinfo/behave
>
>


-- 
陈刚
[e-mail]