SORT BY:

LIST ORDER
THREAD
AUTHOR
SUBJECT


SEARCH

IPS HOME


    [Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

    RE: FCIP: Comment 120



    > >I also believe that the rough
    > > consensus of the WG is not to change FCIP in this fashion.
    > 
    > I am *not* advocating a change.  I am simply concerned that the FCIP 
    > text is neither ruling out certain possibilities explicitly, 
    > nor defining the
    > architectural hooks to cater to all possibilities.  I'd be 
    > perfectly okay with
    > a statement that states that each IP-TCP Address-Port belongs 
    > to exactly
    > one Pair (though the text then should ideally say why).
    
    That's a reasonable request that I'm sure the FCIP draft editor
    would be happy to accommodate.  The reason why is probably going
    to be some version of simplicity.
    
    Thanks,
    --David
    ---------------------------------------------------
    David L. Black, Senior Technologist
    EMC Corporation, 42 South St., Hopkinton, MA  01748
    +1 (508) 249-6449 *NEW*      FAX: +1 (508) 497-8500
    black_david@emc.com         Cell: +1 (978) 394-7754
    ---------------------------------------------------
    
    
    > -----Original Message-----
    > From: Mallikarjun C. [mailto:cbm@rose.hp.com]
    > Sent: Thursday, May 09, 2002 7:32 PM
    > To: Black_David@emc.com; roweber@acm.org; ips@ece.cmu.edu
    > Subject: Re: FCIP: Comment 120
    > 
    > 
    > David,
    > 
    > From the text that Murali posted on this thread - 
    > 
    > " The FC-BB-2_IP Reference Model supports one logical IP Interface and
    > allows sharing a 4-byte IPv4 or 16-byte IPv6 address in the 
    > following ways:
    > 
    > a)A single IP address per FC-BB-2_IP device
    >   - A single IP address shared by all FC/FCIP Entity pairs"
    > 
    > 
    > It's a simple extension to assume that the same well-known TCP port
    > *could* be used for all Pairs, which results in multiple 
    > Pairs at the same *address*
    > necessitating a *name* of some sort to identify the intended 
    > FSF recipient.
    > 
    > >Having seen no other
    > > participants in this discussion,
    > 
    > Please see Murali's message.
    > 
    > >I also believe that the rough
    > > consensus of the WG is not to change FCIP in this fashion.
    > 
    > I am *not* advocating a change.  I am simply concerned that the FCIP 
    > text is neither ruling out certain possibilities explicitly, 
    > nor defining the 
    > architectural hooks to cater to all possibilities.  I'd be 
    > perfectly okay with
    > a statement that states that each IP-TCP Address-Port belongs 
    > to exactly
    > one Pair (though the text then should ideally say why). 
    > 
    > Thanks.
    > --
    > Mallikarjun
    > 
    > Mallikarjun Chadalapaka
    > Networked Storage Architecture
    > Network Storage Solutions Organization
    > Hewlett-Packard MS 5668 
    > Roseville CA 95747
    > cbm@rose.hp.com
    > 
    > 
    > ----- Original Message ----- 
    > From: <Black_David@emc.com>
    > To: <roweber@acm.org>; <ips@ece.cmu.edu>
    > Cc: <cbm@rose.hp.com>
    > Sent: Thursday, May 09, 2002 1:39 PM
    > Subject: RE: FCIP: Comment 120
    > 
    > 
    > > > > Besides, I don't see anything in the current document 
    > that prohibits
    > > multiple
    > > > > FC/FCIP Entity Pairs from sharing the same IP/TCP 
    > address/port - and I
    > > believe
    > > > > that is the right architectural approach.
    > > > 
    > > > In the Orange County meeting, the FCIP contributors were 
    > told that we
    > > > had to stop relying on knowing the exact IP Address and 
    > Port to decide
    > > > the end points of the FCIP Links. We were NOT told that we had to
    > > > allow multiple FC/FCIP Entity Pairs to share a single IP 
    > Address/Port.
    > > > In fact, I believe we were told that we could proceed using such
    > > > an assumption.
    > > 
    > > While Ralph doth protest a bit much, I believe he has the 
    > better part
    > > of this discussion.  It is completely reasonable to identify an IP
    > > service instance as "that which responds when contacting TCP port X
    > > at IP address a.b.c.d" - for example, there's only one http server
    > > behind port 80, but there might be a different one behind port 8080.
    > > 
    > > If there's a NAT involved, the discussion gets a bit subtle, but
    > > the principle remains - one opens a TCP connection to port X at
    > > IP address a.b.c.d, and expects to talk directly to a 
    > specific instance
    > > of an IP-based service.  I don't see any need to change FCIP to
    > > allow multiple FC/FCIP Entity Pairs to share a single IP 
    > Address/Port;
    > > the fact that multiple TCP connections to the same FC/FCIP Entity
    > > Pair can share that TCP Port seems sufficient.  Having seen no other
    > > participants in this discussion, I also believe that the rough
    > > consensus of the WG is not to change FCIP in this fashion.
    > > 
    > > Thanks,
    > > --David
    > > ---------------------------------------------------
    > > David L. Black, Senior Technologist
    > > EMC Corporation, 42 South St., Hopkinton, MA  01748
    > > +1 (508) 249-6449 *NEW*      FAX: +1 (508) 497-8500
    > > black_david@emc.com         Cell: +1 (978) 394-7754
    > > ---------------------------------------------------
    > > 
    > > 
    > > 
    > 
    


Home

Last updated: Fri May 10 09:18:43 2002
10047 messages in chronological order