SORT BY:

LIST ORDER
THREAD
AUTHOR
SUBJECT


SEARCH

IPS HOME


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

    Re: iSCSI Naming and Discovery Requirements Document



    Jim, Julian,
    
    I too have been assuming that a session is the I_T nexus and that
    the initiator and target ends of the session are (in SAM terms)
    service delivery ports.
    
    My question on the Naming and Discovery document was one of
    consistent use of terminology.  Also, as far as I can tell,
    over 90% of the content of the Naming and Discovery document
    falls outside the scope of SAM.
    
    Thanks.
    
    Ralph Weber
    
    julian_satran@il.ibm.com wrote:
    
    >
    >
    > Ralph,
    >
    > I think that Jim has captured well the basics and the session endpoint is
    > the equivalent of the SAM delivery port.
    >
    > Julo
    >
    > "Jim Hafner" <hafner@almaden.ibm.com> on 28/11/2000 19:16:12
    >
    > Please respond to "Jim Hafner" <hafner@almaden.ibm.com>
    >
    > To:   ips@ece.cmu.edu
    > cc:
    > Subject:  Re: iSCSI Naming and Discovery Requirements Document
    >
    > Ralph,
    >
    > IMHO, there is as yet some confusion with respect to the interpretation of
    > SAM's "service delivery port" for iSCSI.  For example, is a service
    > delivery port just a NIC (and so is addressed by MAC), or is it an
    > ipaddress:TCPport (that might get resolved to any one of a number of NICs)
    > or can a service delivery port span multiple ipaddress:TCPport connections
    > (so that it is defined as the coalescing of all the socket connections in a
    > single session)?
    >
    > As you are aware, the answer has implications with respect to reservations
    > and the like, so this needs to be answered carefully.
    >
    > My "sense of the group" is that the third option captures what is desired
    > for reservations.  Namely, the iSCSI session is presented to the SCSI layer
    > as defining the I_T nexus.  This means (I think) that the service delivery
    > port is neither the physical NIC nor the logical ipaddress:TCPport but more
    > precisely the "endpoint" of a session.
    >
    > I could be wrong both on the implications of session in this context and in
    > my "sense of the group", but I throw this out for discussion and
    > clarification.
    >
    > Jim Hafner
    
    
    
    


Home

Last updated: Tue Sep 04 01:06:16 2001
6315 messages in chronological order