SORT BY:

LIST ORDER
THREAD
AUTHOR
SUBJECT


SEARCH

IPS HOME


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

    RE: iSCSI Discovery and SendTargets or Expediency vs. Planning



    
    I'm sensing sarcasm here...  If so, lets take the jokes offline.  People
    have better things to do.  If not, then consider my responses below...
    Nothing is ever "easy to do" as you well know.
    
    > 
    > 
    > Paul,
    > 
    > Just to be sure that I understood well your line of thought:
    > 
    > - solution 1 is good because it is expedient (easy to do!)
    
    I don't believe that was the crux of my message.  While option 1 is
    expedient, it more importantly preserves the utility of the SendTargets
    functionality within the current spec.  As I described, most of what you
    need to support SendTargets will have to be implemented for general iSCSI
    login, authentication and command text processing.  It will be easy for
    customers to make good use of this feature in both small and large
    environments.
    
    > -future extensions should not be considered at this time 
    > (they will become
    > expedient in their own time!)
    
    We should consider future extensions to a point, but we are all looking
    forward to a initial stable version of this spec, and we should make sure we
    are covering the most common and important cases.  Of course this is a
    judgement call.
    
    > -commonality with other discovery protocols has no appeal (you do net
    > mention it)
    > 
    
    Certainly we should not re-invent the wheel, but there is a fine line
    between using a plethora of protocols and associated management to cover a
    space and making a small augmentation to the base protocol to fill a basic
    need.  My opinion is that SendTargets falls on the side of the line to
    include in the iSCSI spec.
    
    > I would like to point out that the SendTargets should not be 
    > considered as
    > a standalone thing.
    > It will (has to) be supported by a management infrastructure that:
    > 
    > - has to install the names
    > -check and invalidate  them as needed
    > 
    > etc.
    > 
    
    Very true, but there is great flexibility in how this is done without
    changing the existing specification or creating a new one.  This may be done
    on the targets themselves, on master targets, independent management servers
    or no-where at all.
    
    > Should we start also adding ReceiveTargets to install the names in the
    > targets (it very easy to add!).
    > And how about certificates, ACLs etc (we could accommodate 
    > those too1).
    > 
    
    Now I'm sure I'm hearing sarcasm...  This is device management.  While we
    are at it, why not consider adding a command to set the baud-rate on the
    target serial port...
    
    Paul
    


Home

Last updated: Tue Sep 04 01:04:33 2001
6315 messages in chronological order