SORT BY:

LIST ORDER
THREAD
AUTHOR
SUBJECT


SEARCH

IPS HOME


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

    Re: iSCSI: remove recovery from transport-layer connection failure(?)



    
    I agree with John that we should keep it.  We had put it in so we
    could recover from lost connections for any SCSI device.  Without
    this mechanism, the only device that would be recoverable is disk.
    
    BTW, NDMP is a file-level protocol; it is used by backup products
    to instruct a file server to write certain files to tape.  We
    still need to be able to write tape at the block level, either
    directly or via a third-party copy.
    
    Even if we end up with a single connection per session, we will
    still have to acknowledge receipt of status to the target, and
    be able to recover sessions if the TCP stream is lost for some
    reason.
    
    --
    Mark
    
    
    John Hufferd/San Jose/IBM wrote:
    > 
    > Costa,
    > This is not a replacement for Tape.  Tape is used for other things besides
    > Backup.  Also, NDMP is a Client/Server  protocol.  Someone still needs to
    > talk to the tape with SCSI or FC.  The actual tape needs to be locatable
    > anywhere. This is still the need for iSCSI.
    > 
    > .
    > .
    > .
    > John L. Hufferd
    > 
    > csapuntz@csapuntz-u1.cisco.com@ece.cmu.edu on 09/22/2000 02:15:06 PM
    > 
    > Sent by:  owner-ips@ece.cmu.edu
    > 
    > To:   ips@ece.cmu.edu
    > cc:
    > Subject:  iSCSI: remove recovery from transport-layer connection failure(?)
    > 
    > Currently, iSCSI is spec'ed to recover from transport-layer
    > connection failures.
    > 
    > The main motivation for this decision was to support tape backup
    > applications that are quite sensitive to any failures that get
    > propogated to their layer.
    > 
    > However, there are other ways of doing backup over IP than
    > SCSI/iSCSI.  NDMP (www.ndmp.org) and devices that implement it are
    > already shipping.
    > 
    > So, perhaps we can remove the requirement of recovering from
    > transport-layer connection failures in iSCSI. This would simplify
    > the protocol somewhat.
    > 
    > Thoughts?
    > 
    > -Costa
    
    -- 
    Mark A. Bakke
    NuSpeed, Inc.
    mark.bakke@nuspeed.com
    763.398.1054
    


Home

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