SORT BY:

LIST ORDER
THREAD
AUTHOR
SUBJECT


SEARCH

IPS HOME


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

    RE: Towards Consensus on TCP Connections



    John,
    
    I would appreciate something a bit more specific.  I was not making a
    statement about existing products bolted directly into the system
    back-plane.  Over a network, less is known.  Again, for your type of
    solution, you would be better off making a tunnel for the reasons mentioned.
    Discuss the details and not your association.
    
    Doug
    
    > -----Original Message-----
    > From: hufferd@us.ibm.com [mailto:hufferd@us.ibm.com]
    > Sent: Friday, August 11, 2000 2:06 PM
    > To: Douglas Otis
    > Cc: ips@ece.cmu.edu; Stephen Byan
    > Subject: RE: Towards Consensus on TCP Connections
    >
    >
    >
    >
    > Doug,
    > I do not think that such contentious statements are useful.
    >
    > If you think that Storage Controllers from EMC and IBM do not
    > scale, then I
    > think there are a number of Companies that  would radically disagree.
    >
    > Again, I am asking that we first focus back on the original intent of Host
    > to Storage Controllers.
    >
    > .
    > .
    > .
    > John L. Hufferd
    >
    > Internet address: hufferd@us.ibm.com
    >
    >
    > "Douglas Otis" <dotis@sanlight.net> on 08/11/2000 01:06:44 PM
    >
    > To:   John Hufferd/San Jose/IBM@IBMUS, <ips@ece.cmu.edu>
    > cc:   "Stephen Byan" <Stephen.Byan@quantum.com>
    > Subject:  RE: Towards Consensus on TCP Connections
    >
    >
    >
    > John,
    >
    > Hard to imagine a Storage Controller unrelated to disk drives.  Are we
    > discussing a protocol or your product spec?  Even the lowly IDE interface
    > started out as a controller.  The quagmire is the mess created by the
    > stateful mid-stream bottleneck you call a controller.  Here you are mixing
    > protocols which you abhor.  You are not using a network or protocol
    > effectively if you refuse to allow this interface to be suitable at the
    > end-point, the device.  The fact that your architecture can not scale
    > speaks
    > loudly.  If you wish to use this interface at a controller, it should
    > assume
    > an identical role to a device.
    >
    > Doug
    >
    
    
    


Home

Last updated: Tue Sep 04 01:07:52 2001
6315 messages in chronological order